REFERENCE LIST Once open in PDF, on the left side of the screen, click on the bookmark button to open bookmarks so you can easily go to the article(s) referenced during various presentations.

Diabetes Impact on Patient Outcomes • Costs of Diabetes Complications: Hospital-based care and absence from work for 392,200 people with Type 2 Diabetes and matched control participants in Sweden • Hospital Readmissions among Commercially Insured and Medicare Advantage Beneficiaries with Diabetes and the Impact of Severe Hypoglycemic and Hyperglycemic Events

Conversations Before the Crisis: DNR Should Not be a Last-Minute Decision • IHI: End-of-Life Care Conversations: Medicare Reimbursement FAQs • CAPC: How to Create a CPAC User Account • www.theconversationproject.org

Heart Failure Update: Pharmacology 2017 ACC/AHA/HFSA Focused Update of the 2013 ACCF/AHA Guideline for the Management of Heart Failure

Men’s Health Update

Ortho Update: When to Refer Common Hand Problems • www.orthobullets.com • The Role of Botulinum Toxin in Vasospastic Disorders of the Hand

Ortho Update: What’s New in Arthritic Joint Replacement? • Outpatient Total Joint Arthroplasty • Identifying the Unicondylar Arthroplasty Patient

Covid Long-Haul Syndrome Long Covid—mechanisms, risk factors, and management

Update on Epic • A Step-by-Step Time-Saving Approach to Coding Office Visits • Medicare Preventive Services Coding • Annual Comprehensive Care Visit form

Medication Affordability

Cardiology Update: Invasive Therapies for Complex Cardiac Conditions • Prevalence and Outcomes of Unoperated Patients with Severe Symptomatic Mitral Regurgitation and Heart Failure • Percutaneous Renal Denervation in patients with Treatment-Resistant Hypertension: final 3-year report of the Symplicity HTN-1 Study

Diabetologia (2020) 63:2582–2594 https://doi.org/10.1007/s00125-020-05277-3

ARTICLE

Costs of diabetes complications: hospital-based care and absence from work for 392,200 people with type 2 diabetes and matched control participants in Sweden

Emelie Andersson1 & Sofie Persson1,2 & Nino Hallén 3 & Åsa Ericsson4 & Desirée Thielke3 & Peter Lindgren 1,5 & Katarina Steen Carlsson1,2 & Johan Jendle6

Received: 22 April 2020 /Accepted: 28 July 2020 / Published online: 24 September 2020 # The Author(s) 2020

Abstract Aims/hypothesis The risk of complications and medical consequences of type 2 diabetes are well known. Hospital costs have been identified as a key driver of total costs in studies of the economic burden of type 2 diabetes. Less evidence has been generated on the impact of individual diabetic complications on the overall societal burden. The objective of this study was to analyse costs of hospital-based healthcare (inpatient and outpatient care) and work absence related to individual macrovascular and microvascular complications of type 2 diabetes in Sweden in 2016. Methods Data for 2016 were retrieved from a Swedish national retrospective observational database cross-linking individual- level data for 1997–2016. The database contained information from population-based health, social insurance and socioeconomic registers for 392,200 people with type 2 diabetes and matched control participants (5:1). Presence of type 2 diabetes and of diabetes complications were derived using all years, 1997–2016. Costs of hospital-based care and of absence from work due to diabetes complications were estimated for the year 2016. Regression analysis was used for comparison with control participants to attribute absence from work to individual complications, and to account for joint presence of complications. Results Use of hospital care for complications was higher in type 2 diabetes compared with control participants in 2016: 26% vs 12% had ≥1 hospital contact; there were 86,104 vs 24,608 outpatient visits per 100,000 people; and there were 9894 vs 2546 inpatient admissions per 100,000 people (all p < 0.001). The corresponding total costs of hospital-based care for complications were €919 vs €232 per person (p < 0.001), and 74.7% of costs were then directly attributed to diabetes (€687 per person). Regression analyses distributed the costs of days absent from work across diabetes complications per se, basic type 2 diabetes effect and unattributed causes. Diabetes complications amounted to €1317 per person in 2016, accounting for possible complex interactions (25% of total costs of days absent). Key drivers of costs were the macrovascular complications angina pectoris, heart failure and stroke; and the microvascular complications eye diseases, including retinopathy, kidney disease and neuropathy. Early mortality in working ages cost an additional €579 per person and medications used in risk-factor treatment amounted to €418 per person. Conclusions/interpretation The economic burden of complications in type 2 diabetes is substantial. Costs of absence from work in this study were found to be greater than of hospital-based care, highlighting the need for considering treatment consequences in a societal perspective in research and policy.

Keywords Costsandcostanalysis .Diabetescomplications .Diabetesmellitus,type2 .Hospitalcosts .Insurance,disability .Sick leave

Emelie Andersson and Sofie Persson are joint first authors. Katarina Abbreviations Steen Carlsson and Johan Jendle are joint senior authors. DRG Diagnosis-related group Electronic supplementary material The online version of this article ESRD End-stage renal disease (https://doi.org/10.1007/s00125-020-05277-3) contains peer-reviewed but ICD International Classification of Diseases unedited supplementary material, which is available to authorised users. IHD Ischaemic heart disease NBHW National Board of Health and Welfare * Katarina Steen Carlsson [email protected] NPDR National Prescribed Drugs Register NPR National Patient Register Extended author information available on the last page of the article Diabetologia (2020) 63:2582–2594 2583

Introduction complications and the extent of the economic burden falling on the individual where insurance coverage is lacking or Type 2 diabetes is a prevalent and costly chronic disease asso- inadequate. ciated with several health complications. Worldwide, the Studies on the cost of illness combine data from a mix of prevalence of diabetes has increased over the last decades sources, often using aggregate costs where a part of the costs is [1] and, in 2019, there were about 463 million adults with attributed to the specific disease. Complications related to type diabetes according to the International Diabetes Federation 2 diabetes are conditions that may also occur in the general [2]. Rising health expenditure in absolute values and as population, such as cardiovascular diseases. Hence, analyses percentage of gross national product, as shown in, for of the costs of type 2 diabetes usually report the total costs instance, Organisation for Economic Cooperation and incurred by people with diabetes and the cost attributed to Development (OECD) health data [3], is a concern for deci- diabetes when applying top-down estimation strategies. For sion makers in national health systems as well as insurance- example, the ADA's report on the economic costs of diabetes based systems. in 2017 attributed 57% of the hospital inpatient care costs The economic burden of diabetes complications is known incurred by people with diabetes to diabetes per se (US$70 to go beyond resource use and costs in the health sector. billion of US$123 billion) [4]. Costs of disease may also be Previous studies on the cost burden of type 2 diabetes have derived from individual-level data with population coverage, demonstrated that costs from reduced productivity (including preferably holding longitudinal information from multiple e.g. work days absent, reduced labour force participation, administrative registers. A Danish study of number of days mortality) constitute a substantial part of the total cost burden, absent from work due to diabetes-related complications as shown by, for instance, the most recent updates on econom- showed that several of the investigated complications led to ic costs of diabetes in the USA from 2017 and in the UK from more than 3 months of work absence for the average person 2012 [4–6]. Less evidence has been generated on the econom- with diabetes [6]. ic impact of individual diabetic complications and little is Recent years have seen an increase in studies using diag- known about which complications are the key drivers of the nostic codes and administrative registers to identify people overall societal burden. The identification of key cost drivers with diabetes and conduct epidemiological, clinical and among complications informs research and policy of where economic research. Some examples include studies on total further actions and tools are needed. Such data support clini- costs of care for people with diabetes [7–10], on prevalence cians, employers, organisations and society with facts on the and incidence of diabetes [11, 12] and on types of glucose- value of preventing and postponing the incidence of diabetes lowering medication and clinical outcomes [13]. Similar 2584 Diabetologia (2020) 63:2582–2594 topics have also been addressed using patient cohorts regis- of diabetes complications, we used the full longitudinal data- tered in quality registers or incidence registers and using base (1997–2016). outcome data from national health registers in Sweden People with diabetes of labour market active ages (16– [14–21]. 70 years) were selected for the database if they, at any point Sweden has a long tradition of reporting population-based during 1997–2016, fulfilled at least one of the following information into registers, and data can be record-linked criteria: between registers for research using the Swedish personal identification number. Register-based definitions of diabetes 1) At least one healthcare visit or inpatient stay with diabetes may rely on diabetes diagnosis solely [12]oruseacombina- as main or secondary diagnosis by the International tion of observed healthcare contacts with diabetes diagnosis Classification of Diseases version 10 (ICD 10; http:// and filled prescriptions of different types of glucose-lowering apps.who.int/classifications/icd10/browse/2016/en) medications [8, 13, 22]. This study used the latter strategy codes for type 1 diabetes (E10), type 2 diabetes (E11) or with the purpose of estimating costs of hospital (inpatient unspecified diabetes (E14) in any year in the NPR from and outpatient) care and costs of absence from work related 1997 to 2016. to individual macrovascular and microvascular complications 2) At least two dispensed prescriptions of glucose-lowering of type 2 diabetes in Sweden. Using cross-linked microdata, medication (Anatomic Therapeutic Chemical (ATC) matched control participants and straight-forward empirical Classification System codes, A10) with dispense dates strategies, this study addresses previously identified research not more than 6 months apart in the NPDR, from 1 gaps in the economic costs of type 2 diabetes and the contri- July 2005 to 31 December 2017. butions of individual diabetes complications. It also provides new insights into the balance of costs between the healthcare Women treated with glucose-lowering medication with a sector and other sectors. coinciding registration of pregnancy were not included in the study population but were eligible again after 2 years. An index date was defined as the first documented occurrence from 1997 and onwards in the NPR, or the first documentation Methods in the NPDR from 1 July 2005 and onwards, whichever occurred first. Data For each person with diabetes, Statistics Sweden matched control participants from the general population (5:1) using The analysis of the cost burden of type 2 diabetes complica- exact year of birth, sex and region of residence in the index tions in 2016 used data from a retrospective observational year. This enabled incremental analysis of costs of hospital study database [22]. The database cross-links 20 years of care and absence from work for people with diabetes individual-level data, 1997–2016, from three national author- compared with control participants. The design allowed ities in Sweden: the National Board of Health and Welfare control participants to switch to the diabetes group at a later (NBHW), Försäkringskassan (the Swedish Social Insurance stage if they then fulfilled inclusion criteria for diabetes. Agency) and Statistics Sweden. The NBHW identified people Further information on the data sources, retrieval process with diabetes and provided data from the National Patient and control selection can be found in the ESM (pages 2–6). Register (NPR) on outpatient and inpatient hospital healthcare The study database includes background data (demograph- [23], and from the National Prescribed Drugs Register ic and socioeconomic), exposure data (diabetes complica- (NPDR) on filled prescriptions of selected drugs [24]. tions) and outcome data (healthcare use, filled prescriptions, Further register description is available in electronic supple- mortality and days absent from work). mentary material (ESM) Table 1. Demographic and socioeco- nomic background characteristics were obtained from Statistics Sweden from the Register of the Total Population Empirical strategies and the LISA database (longitudinal integrated database for health insurance and labour market studies [25]). Selecting the type 2 diabetes population The empirical defi- Försäkringskassan supplied data on days absent from work nition of type 2 diabetes was similar to previous registry-based from the MiDAS database (Micro-Data for the Analysis of diabetes studies [7, 8, 13]. From the study database we omitted Social Insurance) covering schemes of sickness and rehabili- people with type 1 diabetes defined as having inpatient and/or tation benefits, and sickness and activity compensation [26]. outpatient care registered with type 1 diabetes (ICD E10), but This study calculated costs of complications in the year 2016 no registration of type 2 diabetes (E11) or unspecified diabetes based on use of hospital care and absence from work in 2016. (E14), as main or supplementary diagnosis in the NPR during To identify the population with type 2 diabetes and presence the study period. All other cases were included as type 2 Diabetologia (2020) 63:2582–2594 2585 diabetes (see the ESM for more information on selection prescription medications for preventive type 2 diabetes treat- process). ment included glucose-lowering medication, and key preven- tive anti-hypertensive and lipid-lowering medications were Defining diabetes complications Complications were defined estimated by comparison with control participants, but no using ICD codes and procedure codes (Classification of attribution to individual complications was attempted. Healthcare Interventions [Klassifikation av. vårdåtgärder (KVÅ); NBHW], which holds surgical and non-surgical inter- Strategy for costs of absence from work We used information ventions including diagnostic procedures in line with previous from Försäkringskassan on (short-term) sickness and rehabil- studies, as referenced in ESM Table 2). The incidence and itation benefits, and (long-term) sickness and activity compen- prevalence rates of complications were identified through sation, in 2016 [29]. Net days absent from work was derived registrations of main diagnosis in hospital-based care. The using start and stop dates for periods with sick pay and the empirical strategy was not expected to capture all early stages degree of sick leave. The cost of absence from work was of complications, non-emergency needs and problems not valued by multiplying net days with age- and sex-specific necessarily needing assistance from medical care, but to have earnings including payroll taxes [30]fromlabourmarketdata a comprehensive coverage of use of hospital resources. at Statistics Sweden. The analysis included all individuals The characteristics of each complication and its expected <66 years old irrespective of labour market status in 2016. impact on work capacity guided the empirical strategy for Diagnoses recorded at Försäkringskassan are insufficient identifying complications as either an event or a chronic state as a source for attributing spells of sick leave in 2016 to a in the estimation of absence from work. Most complications specific complication of diabetes. Instead, costs of days absent were considered chronic; that is, after the first hospital care from work were attributed to listed diabetes complications contact, they were assumed to remain, with potential impact using regression analyses, including matched control partici- on absence from work for the remainder of the working life. pants and adjusting for level of education (see the Statistical This work used a dichotomous definition (yes/no) of presence analyses section). of each complication and thus covered the range of impact Costs of excess mortality in working ages were calculated through potential progression. Four complications (severe basedonthehumancapitalmethod[30]. Age- and sex- events of hypoglycaemia and hyperglycaemia requiring specific excess mortality in the diabetes group compared with hospital care, ketoacidosis and coma) were considered events the general population were multiplied by discounted age- and as commonly used in health economic evaluations. See also sex-specific net earnings, including pay roll taxes and ESM Linkage of diabetes complications to costs. accounting for mean work force activity up to 65 years of age. Chronic state complications were: eye disease and diabetic retinopathy excluding vision loss and blindness (eye disease Statistical analyses for short); vision loss and blindness; neuropathy; peripheral vascular disease; amputation; kidney disease; end-stage renal Descriptive analyses The descriptive analyses for diabetic and disease (ESRD) with dialysis or kidney transplantation; isch- control participants report demographic and socioeconomic aemic heart disease (IHD); angina pectoris; heart failure; atrial data as number and proportion for categorical variables and as fibrillation; diabetic foot and/or ulcers; and osteoarthritis. mean and standard deviation, or median and interquartile range, Myocardial infarction and stroke allowed both event- and for continuous variables. The univariate analysis used t test for chronic state-related consequences. For these complications, differences between study groups since the central limit theo- an event was defined as an inpatient care episode including rem guarantees near normality of sample means in large-sample potential readmission within 30 days [14, 15]. Outpatient studies like ours [31]. The prevalence rates of complications in healthcare use with main diagnosis of myocardial infarction 2016 were measured as the rate per 100,000 people to illustrate or stroke was considered part of the chronic state of history of long-term burden after first onset of the complication, and were the respective events. compared using standardised differences. The presence of multimorbidity was analysed using cross- Strategy for costing hospital care resource use and medica- tabulation of number of people with diabetes and pairwise combi- tions We used diagnosis-related group (DRG) codes and the nations of studied diabetes complications in 2016 [8]. The results main diagnosis to assign costs to hospital-based care, pricing are presented as a bubble diagram where the size of the bubble all NPR healthcare contacts in 2016 using contact-specific indicates the proportion of people with diabetes having each weights from the Nordic Diagnosis-Related Group complication, and the lines in between complications indicate (NordDRG) nomenclature [27] and the national price for a the number of people with both complications (Fig. 1). DRG weight in 2016 [28]. The proportion attributed to diabe- tes was estimated as the increase in costs per 100 people Cost analyses The total costs of diabetes were calculated by compared with control participants. Incremental costs of summarising costs for hospital-based care, medication and 2586 Diabetologia (2020) 63:2582–2594

Fig. 1 Network of comorbidity interactions for people with type 2 Angina pectoris IHD diabetes in 2016 (N=392,200) Diabetic foot

Peripheral vascular Eye disease and disease diabetic retinopathy Myocardial infarction

Neuropathy

Kidney disease

Atrial fibrillation

Stroke

Heart failure Osteoarthritis

Bubble size corresponds to observed presence of complication in 2016 (incident or prevalent prior to 2016)

10–24.9% >25%

7–9.9% 4–6.9%

Line width corresponds to number of individuals with presence of both complications in 2016 43,000

15,000–25,000 10,000–14,999

4000–9999 absence from work in 2016. Ordinary least squares regression potential differences in effects of K complications on work analysis investigated the correlation between costs of days absence across study groups (whereby K indicates the number absent from work and complication indicators, see Equation of complications). (1). Regression analyses have been applied in several studies Costs of days absent from work of diabetes complications [6, 8, 32, 33]; in our study, the large i sample size allowed the use of simple methods as the analyses ¼ β þ β Education þ β Diabetes 1 2−4 i 2−4 5 i depend on sample means and variances [31]. þ β Complications þ β Diabetes The empirical model distributed costs of days absent from 6−26 i; K 27−42 i work across diabetes complications, accounting for potential  Complications þ εi ð1Þ multiple comorbidities at the individual level, and including i; K education as a confounding factor. It did not include demo- for individual i. The estimated cost of work absence related to graphic or other socioeconomic characteristics as the aim was a complication, for an average individual with diabetes and to obtain an mean value across individuals irrespective of age, the complication, consisted of four parts: sex or other basic characteristics. Equation (1) shows the β þ β þ β þ β ð Þ empirical model, including interaction effects to explore ðÞConstant ðÞDiabetes ðÞComplication ðÞDiabetesÂComplication 2 Diabetologia (2020) 63:2582–2594 2587

Table 1 Demographics of the study population The first term is the regression constant β(Constant),whichis a basic mean cost of absence from work in the estimation Demographic Type 2 diabetes Control sample. It averages costs of days absent from work for all N=392,200 N=1,643,170 individuals, with and without work absence, and with and Index year (median; IQR) 2008 (2005; 2013) 2009 (2005; 2013) without complications etc. The second term β(Diabetes) is the basic diabetes effect, which is not attributed to any specific Control participants available in 2016 complication but differentiates the diabetes group from 5 200,214 (51.0) control participants, including potential differences in the 4 106,126 (27.1) distribution of other diseases not explored in this study. 3 54,781 (14.0) Of primary interest for the analyses was the part of the costs of 2 23,202 (5.9) absence from work that was associated with complications in 16849(1.7) type 2 diabetes, which is the sum of the basic complication effect 01028(0.3) and the interaction effect β Complication + β Diabetes Complication . ( ) ( × ) Year of birth <1940 39,312 (10) 123,819 (7.5) 1940–1949 139,491 (35.6) 541,711 (33) Results 1950–1959 109,537 (27.9) 484,661 (29.5) 1960–1969 62,681 (16) 295,487 (18) – Descriptive analysis 1970 1979 25,786 (6.6) 123,517 (7.5) 1980–1989 11,007 (2.8) 52,559 (3.2) ≥ The study population contained 392,200 people with type 2 1990 4386 (1.1) 21,416 (1.3) diabetes alive and resident in Sweden in 2016, identified in Age in 2016, mean (SD) 62.8 (12.3) 61.5 (12.3) health data registers since 1997 (index years median 2008), with Age in index year, mean (SD) 55.1 (11.5) 54.4 (11.9) 1,643,170 matched control participants. The mean age was Sex 63 years in 2016, with 59% men and 41% women (Table 1). Men 230,539 (58.8) 931,598 (56.7) Panel attrition after the index year was mainly due to diabetes Women 161,661 (41.2) 711,572 (43.3) onset in individuals originally selected as control participants Education (5:1). It implied that while 51% of people with diabetes still Compulsory 117,989 (30.1) 351,786 (21.4) had five control participants in 2016, 0.3% had no control partic- Upper secondary 182,173 (46.5) 738,437 (44.9) ipants and 7.7% had only one or two control participants. This University 86,825 (22.1) 541,910 (33) attrition led to small differences in age and sex distribution Missing 5213 (1.3) 11,037 (0.7) between type 2 diabetes and control participants. There was Data are number (proportion) unless otherwise stated generally a lower achieved level of education among people with diabetes compared with control participants (compulsory education only 30% vs 21%, university 22% vs 33%). These current or previous myocardial infarction (23,300; 6%), angi- data confirm an expected difference in the distribution of socio- na pectoris (22,900; 6%) and eye disease (19,800; 5%). economic characteristics between diabetic and control partici- People with myocardial infarction also had eye disease pants, motivating an inclusion of the level of education in (17,400; 4%) and angina pectoris (16,300; 4%). Heart failure regression analysis to account for these differences. (n = 20,100; 5.1%) showed the most complexity with multiple Table 2 shows a higher morbidity among people with type complications: more than 4000 individuals had heart failure 2 diabetes. The prevalence rates for ten out of 15 complica- and at least one other of six studied complications (eye tions were higher in the diabetes group based on standardised disease, kidney disease, IHD, angina pectoris, myocardial differences, even though mean values were different in all infarction and atrial fibrillation). Fewer interactions were complications. The largest standardised differences were found for the major cardiovascular complications of stroke found for kidney disease, eye disease and angina pectoris. (n = 23,500; 2.3%; where 5300 also had eye disease) and Figure 1 shows the diabetes-related comorbidity by myocardial infarction (n = 33,900; 8.6%; where 8300 also describing the most prevalent diabetes complications and had kidney disease). pairwise links of the most frequent interactions. Eye disease (43%) and kidney disease (16%) had the highest prevalence, and 43,000 individuals (11%) had both of these complica- Costs of diabetes complications in 2016 tions. The second most common interaction (n = 24,000; 6%) was between angina pectoris (11%) and eye disease. People with type 2 diabetes and complications had high hospi- Other frequent interactions were IHD in combination with tal care consumption and absence from work in 2016 2588 Diabetologia (2020) 63:2582–2594

Table 2 Prevalence of diabetes complications in type 2 diabetic Diabetes complication and concomitant conditions Rate per 100,000 individuals p value Std diff (N = 392,200) and control partici- pants (N = 1,643,170) 2016: Type 2 diabetes Control number of individuals with complication per 100,000 in Macrovascular complication population IHD 9619 3288 <0.001 –0.260 Angina pectoris 10,977 3733 <0.001 –0.280 Acute myocardial infarction 8640 3058 <0.001 –0.240 Stroke 5994 2504 <0.001 –0.174 Heart failure 5118 1430 <0.001 –0.208 Atrial fibrillation 7145 4261 <0.001 –0.086 Other sudden death, cause unknown 1 0 0.0277 –0.004 Microvascular complication Eye disease and diabetic retinopathya 43,149 28,295 <0.001 –0.337 Vision loss or blindness 374 211 <0.001 –0.039 Lower extremity disease Neuropathy 6167 3316 <0.001 –0.162 Peripheral vascular disease 4479 764 <0.001 –0.190 Amputation 797 34 <0.001 –0.093 Kidney disease 16,357 2260 <0.001 –0.465 ESRDb 1391 316 <0.001 –0.129 Other complications and events Hypoglycaemia 707 5 <0.001 –0.117 Hyperglycaemia 619 3 <0.001 –0.118 Ketoacidosis 143 0 <0.001 –0.061 Coma 92 13 <0.001 –0.036 Diabetic foot and ulcer 4874 1445 <0.001 –0.182 Osteoarthritis 4001 544 <0.001 –0.158

a Not vision loss and blindness b ESRD with dialysis or kidney transplantation Std diff, standardised difference

(Table 3). The total costs of hospital-based care were €360 the total costs of hospital-based care and absence from work million, of which 74.7% (€269.3 million; €687 per person related to diabetes complications in 2016 were €785.7 million, with diabetes) were attributed to type 2 diabetes from compar- corresponding to €2003 per person with diabetes. IHD condi- ison with corresponding mean costs among control partici- tions angina pectoris and myocardial infarction generated pants. Table 3 presents two alternative estimates of the total absence from work, while the regression analysis costs for days absent from work for people with complica- (Table 3) did not attribute additional days absent to tions. By the high-level analysis accounting for having at least remaining IHD conditions. A similar pattern was seen one recorded complication event and state, respectively, for the event of stroke, although the state of stroke absence from work amounted to €1744 million, of which caused additional absence from work indicating long- 51% (€884 million; €2254 per person with diabetes) was term effects. See ESM Table 3 for more details on attributed to complications (total of event and chronic state). estimations of costs related to absence from work. The total costs of hospital-based care and of absence from Two additional cost items from the data were not distribut- work related to diabetes complications in 2016 were €1153 ed across complications but are included as points of refer- million (€2943 per person with diabetes). ence. First, we observed 1100 additional deaths before age The analysis of costs of absence from work by individual 66 years compared with general age-related mortality. The complications (Table 3; Detailed analysis—by individual loss of working-age years was valued at €227 million from complication) gave a higher estimate of the total costs of lost production (€579 per person). Second, insulin and other absence from work (€2080 million), but the costs attributed glucose-lowering medications cost €135 million in 2016. to the diabetes complications were lower (25%; €516.4 People with type 2 diabetes had 924,902 registered prescrip- million, €1317 per person with diabetes). By this estimate, tions of other risk-factor medications, including drugs used for Diabetologia (2020) 63:2582–2594 2589

Table 3 Diabetes complications and costs in euros in 2016: hospital-based care costs and total costs of days absent from work in 2016

Diabetes complication and concomitant conditions Hospital-based care costs (inpatient admissions Costs of absence from work due to diabetes and outpatient visits) due to diabetes complications complications

Total cost of Attributed to diabetes Total costs of days Attributed to diabetes hospital-based care (€) complication (%)a absent from work (€) complication (%)b

High-level analysis—complications aggregated Hospital-based care (admissions and visits) 360,335,000 75 Absence from work by regression analysis Event 448,026,000 21 State 1,296,035,000 61 Detailed analysis—by individual complication Macrovascular complications IHD 13,549,000 71 100,425,000 0 Angina pectoris 22,992,000 72 146,949,000 31 Acute myocardial infarction 32,248,000 68 Event 12,588,000 11 State 102,118,000 11 Stroke 27,529,000 58 Event 4,812,000 0 State 119,867,000 59 Heart failure 34,269,000 80 74,623,000 43 Atrial fibrillation 14,586,000 33 55,305,000 5 Other sudden death, cause unknown 1439 31 Microvascular complications Eye disease and diabetic retinopathyc 53,187,000 57 654,577,000 16 Vision loss or blindness 108,000 59 11,558,000 48 Lower extremity disease Neuropathy 2,992,000 51 169,134,000 33 Peripheral vascular disease 20,383,000 87 58,233,000 30 Amputation 7,493,000 95 11,410,000 30 Kidney disease 44,526,000 85 307,539,000 24 ESRDd 46,531,000 89 50,858,000 51 Other complications and events Hypoglycaemia 5,606,000 99 15,123,000 21 Hyperglycaemia 2,379,000 100 19,605,000 38 Ketoacidosis 2,947,000 100 4,168,000 22 Coma 951,000 96 3,071,000 51 Diabetic foot and ulcer 12,052,000 91 115,173,000 36 Osteoarthritis 16,005,000 86 43,217,000 34 a Results from Table 5 columns (1) and (2): percentage calculated as [(1)−(2)]/(1). See also ESM Table 4 b Results from regression analysis in ESM Table 3 columns (1) and (2): percentage calculated as (2)/(1) c Not vision loss and blindness d ESRD with dialysis or kidney transplantation the treatment or prevention of hypertension, dyslipidaemia, Combining results in Tables 3 and 4 shows that common eye disease and neuropathy, costing a total of €45 million of complications may induce high total costs, although cost per which €29 million (65%) was a diabetes increment (ESM person is low. People with type 2 diabetes used more hospital Table 4). The total added costs for diabetes related medica- care as a group: 86,104 vs 24,608 per 100,000 outpatient visits tions were €164 million (€418 per person). and 9894 vs 2546 per 100,000 inpatient admissions (both 2590 Diabetologia (2020) 63:2582–2594

Table 4 Number of outpatient visits and inpatient admissions Diabetes complication and Outpatient visits Inpatient admissions per 100,000 people for the popu- concomitant conditions lation of type 2 diabetic (N = Type 2 Control p Type 2 Control p 392,200) and control participants diabetes participants value diabetes participants value (N = 1,643,170) in 2016 Any event (hospital visit, 86,104 24,608 <0.001 9894 2546 <0.001 admission) Macrovascular IHD 2450 813 <0.001 371 99 <0.001 Angina pectoris 1183 396 <0.001 859 245 <0.001 Acute myocardial 208 79 <0.001 963 316 <0.001 infarction Stroke 327 142 <0.001 753 333 <0.001 Heart failure 2482 656 <0.001 1588 291 <0.001 Atrial fibrillation 2868 2409 <0.001 847 578 <0.001 Other sudden death, cause 1 0 0.014 0 0 0.625 unknown Microvascular Eye disease and diabetic 36,112 16,776 <0.001 350 151 <0.001 retinopathya Vision loss or blindness 93 35 <0.001 3 1 <0.001 Lower extremity disease Neuropathy 1295 689 <0.001 25 8 <0.001 Peripheral vascular 1932 292 <0.001 623 90 <0.001 disease Amputation 42 1 <0.001 170 9 <0.001 Kidney disease 9381 2015 <0.001 1450 307 <0.001 ESRDb 22,899 183 <0.001 347 28 <0.001 Other complications and events Hypoglycaemia 973 5 <0.001 270 2 <0.001 Hyperglycaemia 656 3 <0.001 152 0 <0.001 Ketoacidosis 97 0 <0.001 159 0 <0.001 Coma 65 12 <0.001 47 3 <0.001 Diabetic foot and ulcer 2795 278 <0.001 412 42 <0.001 Osteoarthritis 247 42 <0.001 503 71 <0.001

a Not vision loss and blindness b ESRD with dialysis or kidney transplantation p < 0.001) (Table 4). Moreover, 26% of people with type 2 Inpatient admissions related to complications were key diabetes had at least one hospital contact compared with 12% drivers. Two complications also generated sizable outpatient for control participants (p < 0.001; ESM Table 5). More than costs: eye disease and ESRD (Table 5). half of these (15%) had ≥1 registration of eye disease, accounting altogether for 1372 inpatient admissions and 141,631 outpatient visits. The largest differences between Discussion type 2 diabetic and control participants were found among complications common in diabetes, such as kidney disease This study points to substantial costs of complications in and diabetic foot and ulcers. Heart failure and kidney disease hospital care for type 2 diabetes: the incremental costs of generated most inpatient admissions among people with type hospital-based care, about 75% of the total costs of complica- 2 diabetes. Atrial fibrillation had the highest rate among tions, corresponded to €687 per person with diabetes. Costs control participants (578 per 100,000 control participants) outside the health sector were even higher in a societal but this was still a lower level than in type 2 diabetes (847 perspective. With two alternative specifications of the analy- per 100,000). Total costs of hospital-based care were estimat- sis, costs of workdays absent attributed to diabetes complica- ed at €919 per person for the diabetes population and at €232 tions were estimated at €1317 or €2254 per person, respec- for the control group (Table 5 shows results per 100 people). tively. By both specifications, costs of absence from work Diabetologia (2020) 63:2582–2594 2591

Table 5 Hospital care costs in euros per 100 people for participants with type 2 diabetes and control participants in 2016

Diabetes complication and Hospital care costs (€) Costs of outpatient visits (€) Costs of inpatient admissions (€) concomitant conditions (1) (2) (3) (4) (5) (6) (7) (8) (9) Type 2 Control p value Type 2 diabetes Control p value Type 2 diabetes Control p value diabetes

Any hospital care utilisation 91,875 23,222 <0.001 32,399 9104 <0.0001 59,477 14,118 <0.001 Macrovascular complications IHD 3453 982 <0.001 1045 338 <0.001 2408 644 <0.001 Angina pectoris 5860 1668 <0.001 771 264 <0.001 5090 1404 <0.001 Acute myocardial infarction 8226 2587 <0.001 95 32 <0.001 8131 2555 <0.001 Stroke 7022 2935 <0.001 148 63 <0.001 6874 2872 <0.001 Heart failure 8732 1711 <0.001 982 275 <0.001 7750 1436 <0.001 Atrial fibrillation 3717 2492 <0.001 1045 898 <0.001 2682 1594 <0.001 Other sudden death, cause unknown <0.1 <0.1 0.749 <0.1 <0.1 0.014 <0.1 <0.1 0.625 Microvascular complications Eye disease and diabetic retinopathya 13,558 5808 <0.001 12,270 5258 <0.001 1299 539 <0.001 Vision loss or blindness 32 11 <0.001 21 11 <0.001 11 <0.1 <0.001 Lower extremity disease Neuropathy 760 370 <0.001 623 338 <0.001 137 42 <0.001 Peripheral vascular disease 5195 686 <0.001 697 106 <0.001 4498 581 <0.001 Amputation 1911 106 <0.001 63 <0.1 <0.001 1848 106 <0.001 Kidney disease 11,351 1668 <0.001 3653 412 <0.001 7698 1267 <0.001 ESRDb 11,869 1320 <0.001 9630 1024 <0.001 2239 296 <0.001 Other complications and events Hypoglycaemia 1425 11 <0.001 296 <0.1 <0.001 1130 11 <0.001 Hyperglycaemia 602 <0.1 <0.001 106 <0.1 <0.001 507 <0.1 <0.001 Ketoacidosis 750 <0.1 <0.001 <0.1 <0.1 <0.001 750 <0.1 <0.001 Coma 243 11 <0.001 11 <0.1 <0.001 232 11 <0.001 Diabetic foot and ulcer 3073 285 <0.001 887 84 <0.001 2186 211 <0.001 Osteoarthritis 4076 581 <0.001 74 11 <0.001 4013 570 <0.001 a Not vision loss and blindness b ESRD with dialysis or kidney transplantation were higher than costs of hospital-based care, indicating a account for dialysis or kidney transplantation in kidney significant contribution to overall societal burden of compli- disease, but do not allow for further categorising. The esti- cations. Production loss related to excess mortality among mates provide an indication of the mean value of potentially people with type 2 diabetes of labour market active ages added heterogeneous groups. ‘Other complications’ was a mixed further to total costs. group where the chronic conditions diabetic foot and ulcers Detailed analyses of costs of complications indicated that and osteoarthritis were key drivers of societal costs. The the levels of costs of hospital-based care and absence from results point at a remaining unmet need for risk-factor treat- work for macrovascular complications were €96 million and ment as diabetes complications continue to generate substan- €164 million, respectively. Microvascular complications had tial costs of hospital-based care and absence from work. higher costs for hospital-based care (€136 million), and costs A further observation was that the percentage of total costs of absence from work were nearly twice as large as those for of hospital-based care attributed to type 2 diabetes and the macrovascular complications (€287 million). The data percentage of total costs of work absence attributed to diabetes presented underline the fact that common complications may complications varied between complications (Table 3). For add up to high societal costs even though the cost per person is instance, the regression analysis did not attribute any of the low. Two examples of this are the observed costs for eye costs of work absence to the IHD complication, while 71% of disease and kidney disease. Data allowed for separating of hospital costs were attributed to type 2 diabetes. Figure 1 severe vision loss from earlier stages of eye disease and shows that comorbidities were common among people with 2592 Diabetologia (2020) 63:2582–2594

IHD. Both angina pectoris and myocardial infarctions were sickness and activity compensation scheme available for people common in people with IHD and were associated with consid- with permanent loss of work capacity. erable costs of work absence attributed to the complication. The study has limitations. It lacks national data on use of Previous estimates of the economic burden of diabetes primary care related to diabetes complications and information differ from this study in objectives and design but provide about reduced productivity at work (‘presenteeism’)dueto relevant points of reference. A British study from 2012 esti- diabetes complications. Furthermore, the study design did not mated current and future costs of type 1 and type 2 diabetes, cover all people with diabetes but only those with diabetes indi- respectively, using aggregated data and assumptions from the cation in the working-age population. However, state of compli- literature. The results showed that diabetes accounted for 10% cation and use of hospital-based care were captured up to age of total health resource expenditure in 2010/2011 in the UK, 79 years. The Swedish National Diabetes Register (NDR), with where costs of production loss exceed direct healthcare costs an estimated coverage of 91.5%, reported that about 363,000 [5]. The investigators reported total costs and did not compare individuals with type 2 diabetes were registered in 2016 [34]. them with corresponding costs in the general population. On This indicates that, while the study design does not include the other hand, the diabetes increment was estimated in the US people aged 80 years and older, the study population of study by the ADA from 2017, which reported that people with 392,200 individuals still represents most of all cases of type 2 diabetes had 2.3 times higher medical expenditures compared diabetes in Sweden. Moreover, the incremental costs of diabetes with people with no diabetes [4]. This study included both compared with the general population diminish with age [35]. type 1 and type 2 diabetes and did not report costs by individ- Therefore, underestimation of the costs of complications in ual diabetes complications. A German study from 2018 Sweden from onset of type 2 diabetes beyond study inclusion explored the economic impact of diabetes complications and criteria was limited. their interactions using nation-wide statutory health insurance data. The results showed that an increased number of compli- cations was associated with higher total costs [7, 8]. The Conclusion German study, as well as two recent studies from Sweden [9, 10], reported total healthcare costs for people with type 2 This study shows that the burden of complications in type 2 diabetes but did not account for the fact that several compli- diabetes is substantial, and costs due to absence from work are cations are increasingly prevalent with age in the general even greater than costs of hospital-based healthcare. This population. Using the estimated costs of complications among poses a challenge to society in terms of loss of work capacity, control participants, we show that the fraction attributed to and more research and emphasis on guideline implementation diabetes varies substantially between complications (Table 3). are needed in the field of preventive treatment of type 2 diabe- The presence of diabetes complications was sourced from tes complications. 20 years of individual-level longitudinal NPR data capturing complications serious enough to require outpatient physician Data availability Subject-level data from national registers hosted by care or an inpatient admission. Compared with studies with more Swedish national authorities are available for research after formal eval- limited follow-up [6, 7], this study is expected to provide a more uation of the research protocol by the Swedish Ethical Review Authority and by the respective national authorities providing data. Permission to comprehensive perspective on the long-term burden of diabetes conduct research is granted on a case-by-case basis to a limited number of complications. The record-linking between health data and social named people. insurance registers for people with type 2 diabetes and their matched control participants provided a foundation for analyses Funding Open access funding provided by Lund University. Funding for this study was provided by Novo Nordisk A/S, Copenhagen, Denmark to of absence from work related to diabetes complications, which the Swedish Institute for Health Economics, Lund, Sweden. makes results more robust as they do not rely on recall in surveys or combination of results from many different sources as, for Authors’ relationships and activities Four authors are employees of the example, in the cited US and UK studies [4, 5]. Addressing the Swedish Institute for Health Economics (EA, SP, PL, KSC), and three authors are employees of Novo Nordisk (NH, ÅE, DT). The authors had impact of all complications at the same time implied that our the final responsibility for submitting the manuscript. results were less vulnerable to double counting, especially considering multiple comorbidities. The study design also Contribution statement All authors confirm that they meet the implied that we avoided double counting of burden of diabetes International Committee of Medical Journal Editors (ICJME) uniform requirements for authorship and that they have contributed to: design complications since we used the main diagnosis in hospital care. and collection of data; or critical analysis and interpretation of the data; Still, this was a conservative approach as we might underesti- and drafting/critically revising the article. All authors share the final mate the costs since diabetes complications registered as second- responsibility for the content of the manuscript and the decision to submit ary diagnosis with diabetes as main diagnosis were not included. it for publication. Material preparation, data collection and analysis were performed by EA, SP and KSC, as was the writing of the first draft of the A further strength of the analysis was that it also captured the manuscript. KSC and JJ are joint guarantors of this work. long-term consequences of complications through data from the Diabetologia (2020) 63:2582–2594 2593

Open Access This article is licensed under a Creative Commons 12. Tamayo T, Brinks R, Hoyer A, Kuß OS, Rathmann W (2016) The Attribution 4.0 International License, which permits use, sharing, adap- prevalence and incidence of diabetes in Germany. Dtsch Arztebl Int tation, distribution and reproduction in any medium or format, as long as 113(11):177–182. https://doi.org/10.3238/arztebl.2016.0177 you give appropriate credit to the original author(s) and the source, 13. Nyström T, Bodegard J, Nathanson D, Thuresson M, Norhammar provide a link to the Creative Commons licence, and indicate if changes A, Eriksson JW (2017) Novel oral glucose-lowering drugs are asso- were made. The images or other third party material in this article are ciated with lower risk of all-cause mortality, cardiovascular events included in the article's Creative Commons licence, unless indicated and severe hypoglycaemia compared with insulin in patients with otherwise in a credit line to the material. If material is not included in type 2 diabetes. Diabetes Obes Metab 19(6):831–841. https://doi. the article's Creative Commons licence and your intended use is not org/10.1111/dom.12889 permitted by statutory regulation or exceeds the permitted use, you will 14. Adamsson Eryd S, Svensson AM, Franzén S, Eliasson B, Nilsson need to obtain permission directly from the copyright holder. To view a PM, Gudbjörnsdottir S (2017) Risk of future microvascular and copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. macrovascular disease in people with type 1 diabetes of very long duration: a national study with 10-year follow-up. Diabet Med 34(3):411–418. https://doi.org/10.1111/dme.13266 15. Kiadaliri AA, Gerdtham U-G, Nilsson P, Eliasson B, Gudbjörnsdottir S, Steen Carlsson K (2013) Towards renewed health economic simulation of type 2 diabetes: risk equations for References first and second cardiovascular events from Swedish register data. PLoS One 8(5):e62650. https://doi.org/10.1371/journal.pone. 1. NCD Risk Factor Collaboration (NCD-RisC) (2016) Worldwide 0062650 trends in diabetes since 1980: a pooled analysis of 751 16. Persson S, Gerdtham UG, Steen Carlsson K, Swedish Childhood population-based studies with 4.4 million participants. Lancet Diabetes Study Group (2016) Labor market consequences of child- 387(10027):1513–1530. https://doi.org/10.1016/S0140-6736(16) hood onset type 1 diabetes. Econ Hum Biol 23:180– 192. https:// 00618-8 doi.org/10.1016/j.ehb.2016.09.003 2. International Diabetes Federation (IDF) (2019) IDF Diabetes Atlas 17. Steineck I, Cederholm J, Eliasson B et al (2015) Insulin pump 9th edition. Available from http://www.diabetesatlas.org/en/. therapy, multiple daily injections, and cardiovascular mortality in Accessed 19 Nov 2019 18,168 people with type 1 diabetes: observational study. BMJ 350: 3. Organisation for Economic Co-operation and Development h3234. https://doi.org/10.1136/bmj.h3234 (OECD) (2019) OECD Health Statistics 2019. Available from 18. Toresson Grip E, Svensson A-M, Miftaraj M et al (2019) Real- http://www.oecd.org/els/health-systems/health-data.htm. Accessed world costs of continuous insulin pump therapy and multiple daily 4 Dec 2019 injections for type 1 diabetes: a population-based and propensity- 4. American Diabetes Association (2018) Economic costs of diabetes matched cohort from the Swedish National Diabetes Register. – in the U.S. in 2017. Diabetes Care 41(5):917–928. https://doi.org/ Diabetes Care 42(4):545 552. https://doi.org/10.2337/dc18-1850 10.2337/dci18-0007 19. Persson S, Dahlquist G, Gerdtham U-G, Steen Carlsson K (2018) Why childhood-onset type 1 diabetes impacts labour market 5. Hex N, Bartlett C, Wright D, Taylor M, Varley D (2012) Estimating outcomes: a mediation analysis. Diabetologia 61(2):342–353. the current and future costs of type 1 and type 2 diabetes in the UK, https://doi.org/10.1007/s00125-017-4472-3 including direct health costs and indirect societal and productivity 20. Persson S, Dahlquist G, Gerdtham U-G, Steen Carlsson K (2013) costs. Diabet Med 29(7):855–862. https://doi.org/10.1111/j.1464- 5491.2012.03698.x Impact of childhood-onset type 1 diabetes on schooling: a population-based register study. Diabetologia 56(6):1254–1262. 6. Sørensen J, Ploug UJ (2013) The cost of diabetes-related compli- https://doi.org/10.1007/s00125-013-2870-8 cations: registry-based analysis of days absent from work. Econ Res 21. Steen Carlsson K, Landin-Olsson M, Nystrom L et al (2010) Long- Int 2013:1–8. https://doi.org/10.1155/2013/618039 term detrimental consequences of the onset of type 1 diabetes on 7. Kähm K, Laxy M, Schneider U, Holle R (2019) Exploring different annual earnings—evidence from annual registry data in 1990– strategies of assessing the economic impact of multiple diabetes- 2005. Diabetologia 53(6):1084–1092. https://doi.org/10.1007/ associated complications and their interactions: a large claims- s00125-009-1625-z – based study in Germany. Pharmacoeconomics 37(1):63 74. 22. Persson S, Johansen P, Andersson E et al (2020) Days absent from https://doi.org/10.1007/s40273-018-0699-1 work as a result of complications associated with type 2 diabetes: 8. Kähm K, Laxy M, Schneider U, Rogowski WH, Lhachimi SK, evidence from 20 years of linked national registry data in Sweden. Holle R (2018) Health care costs associated with incident compli- Diabetes Obes Metab 22(9):1586–1597. https://doi.org/10.1111/ cations in patients with type 2 diabetes in Germany. Diabetes Care dom.14070 – 41(5):971 978. https://doi.org/10.2337/dc17-1763 23. National Board of Health and Welfare (NBHW) (2019) The 9. Nathanson D, Sabale U, Eriksson JW et al (2018) Healthcare cost National Patient Register. Available from http://www. development in a type 2 diabetes patient population on glucose- socialstyrelsen.se/en/statistics-and-data/registers/register- lowering drug treatment: a nationwide observational study 2006– information/the-national-patient-register/. Accessed 4 Nov 2019 2014. PharmacoEcon - Open 2(4):393–402. https://doi.org/10. 24. National Board of Health and Welfare (NBHW) (2019) 1007/s41669-017-0063-y Läkemedelsregistret [in Swedish; National Prescribed Drugs 10. Sabale U, Bodegard J, Sundstrom J et al (2015) Healthcare utiliza- Register]. Available from http://www.socialstyrelsen.se/statistik- tion and costs following newly diagnosed type-2 diabetes in och-data/register/alla-register/lakemedelsregistret/. Accessed 4 Sweden: a follow-up of 38,956 patients in a clinical practice setting. Nov 2019 Prim Care Diabetes 9(5):330–337. https://doi.org/10.1016/j.pcd. 25. Statistics Sweden (2019) Longitudinal integrated database for 2015.01.001 health insurance and labour market studies (LISA). Available from 11. Jansson S, Fall K, Brus O et al (2015) Prevalence and incidence of http://www.scb.se/en/services/guidance-for-researchers-and- diabetes mellitus: a nationwide population-based pharmaco-epide- universities/vilka-mikrodata-finns/longitudinella-register/ miological study in Sweden. Diabet Med 32(10):1319–1328. longitudinal-integrated-database-for-health-insurance-and-labour- https://doi.org/10.1111/dme.12716 market-studies-lisa/. Accessed 4 Nov 2019 2594 Diabetologia (2020) 63:2582–2594

26. Försäkringskassan (the Swedish Social Insurance Agency) (2019) 30. Drummond MF, Sculpher MJ, Claxton K, Stoddart GL, Torrance MicroData for Analysis of Social Insurance (MiDAS) [in Swedish]. GW (2015) Methods for the economic evaluation of health care Available from http://www.forsakringskassan.se/statistik/kontakta- programmes. Oxford University Press, Oxford statistikenheten. Accessed 22 Nov 2019 31. Mihaylova B, Briggs A, O'Hagan A, Thompson S (2011) Review 27. National Board of Health and Welfare (NBHW) (2016) Vägledning of statistical methods for analysing healthcare resources and costs. till NordDRG—Svensk version 2016 [in Swedish: Instructions for Health Econ 20(8):897–916. https://doi.org/10.1002/hec.1653 NordDRG—Swedish version 2016]. Available from http://www. 32. Gerdtham U-G, Clarke P, Hayes A, Gudbjornsdottir S (2009) socialstyrelsen.se/utveckla-verksamhet/e-halsa/klassificering-och- Estimating the cost of diabetes mellitus-related events from inpa- koder/drg/. Accessed 22 Nov 2019 tient admissions in Sweden using administrative hospitalization 28. National Board of Health and Welfare (NBHW) (2019) data. Pharmacoeconomics 27(1):81–90. https://doi.org/10.2165/ Referensvikter för slutenvårds-och öppenvårdsgrupper, 00019053-200927010-00008 NordDRG 2016 [in Swedish; Reference weights for inpatient and 33. Shao H, Yang S, Fonseca V, Stoecker C, Shi L (2019) Estimating outpatient groups, NordDRG 2016]. Available from http://www. quality of life decrements due to diabetes complications in the socialstyrelsen.se/utveckla-verksamhet/e-halsa/klassificering-och- United States: the Health Utility Index (HUI) diabetes complication koder/drg/. Accessed 22 Nov 2019 equation. Pharmacoeconomics 37(7):921–929. https://doi.org/10. 29. Försäkringskassan (the Swedish Social Insurance Agency) (2020) 1007/s40273-019-00775-8 Sickness compensation. Available from http://www. 34. Swedish National Diabetes Register (NDR) (2017) Nationella forsakringskassan.se/myndigheter_old/arbetsformedlingen/ Diabetesregistret—Årsrapport 2016 [in Swedish; Swedish forsakring/!ut/p/z1/hZBBT8MwDIV_Ta- National Diabetes Register—Annual Report 2016 ]. Available from xV9SRcuskYEVCmrTDSi4oZW4arUmm http://www.ndr.nu/#/arsrapport. Accessed 5 Feb 2020 JFD49xjEBQk6n_zk7z1ZDxR0oLx-s0ZnG7yeWD- 35. Wirehn A-B, Andersson A, Östgren C-J, Carstensen J (2008) Age- p9bMs8XGzWTUo9yixrbf17V1d3ldYwuEbwH-mQVDL_j15zl specific direct healthcare costs attributable to diabetes in a Swedish BL2K66vgC0qwtAWf0AC48-gBpOwn34ozUjZYpCx population: a register-based analysis. Diabet Med 25(6):732–737. 55yGkJ0dJysN-QFi6RPkYWINOreTpZJlon4kLP https://doi.org/10.1111/j.1464-5491.2008.02444.x n9au0xvdX0oCKNFDkqNfIXY45n9NNgQXO8 yxMCGYi8RJcgX9ZxpAydL9JOLvuvbU7d5Cp-QR3noxh/dz /d5/L2dJQSEvUUt3QS80TmxFL1o2XzgyM Publisher’snote Springer Nature remains neutral with regard to jurisdic- E1CQjFBMDhTMDgwSTlIOUVGOTJHUDU3/. Accessed 15 tional claims in published maps and institutional affiliations. June 2020

Affiliations

Emelie Andersson1 & Sofie Persson1,2 & Nino Hallén 3 & Åsa Ericsson4 & Desirée Thielke3 & Peter Lindgren1,5 & Katarina Steen Carlsson1,2 & Johan Jendle6

1 The Swedish Institute for Health Economics, Lund, Sweden 5 Department of Learning, Informatics, Management and Ethics, Karolinska Institutet, Stockholm, Sweden 2 Health Economics Unit, Department of Clinical Sciences, Malmö, Lund University, Lund, Sweden 6 Diabetes, Endocrinology and Metabolism Research Centre, Institute of Medical Sciences, Örebro University, Örebro, Sweden 3 Novo Nordisk A/S, Copenhagen, Denmark

4 Novo Nordisk Scandinavia, Malmö, Sweden Hospital Readmissions among Commercially Insured and Medicare Advantage Beneficiaries with Diabetes and the Impact of Severe Hypoglycemic and Hyperglycemic Events Rozalina G. McCoy, MD MS1,2,3, Kasia J. Lipska, MD MHS4, Jeph Herrin, PhD5,6, Molly M. Jeffery, PhD2,3, Harlan M. Krumholz, MD SM5,7,8,9, and Nilay D. Shah, PhD2,3,10

1Division of Primary Care Internal Medicine, Department of Medicine, Mayo Clinic, Rochester, MN, USA; 2Division of Health Care Policy & Research, Department of Health Sciences Research, Mayo Clinic, Rochester, MN, USA; 3Mayo Clinic Robert D. and Patricia E. Kern Center for the Science of Health Care Delivery, Rochester, MN, USA; 4Section of Endocrinology, Department of Internal Medicine, Yale School of Medicine, New Haven, CT, USA; 5Section of Cardiovascular Medicine, Department of Internal Medicine, Yale School of Medicine, New Haven, CT, USA; 6Health Research and Educational Trust, Chicago, IL, USA; 7Robert Wood Johnson Foundation Clinical Scholars Program, Department of Internal Medicine, Yale School of Medicine, New Haven, CT, USA; 8Center for Outcomes Research and Evaluation, Yale New Haven Health, New Haven, CT, USA; 9Department of Health Policy and Management, Yale School of Public Health, New Haven, CT, USA; 10OptumLabs, Cambridge, MA, USA.

BACKGROUND: Hospital readmission is common among CONCLUSIONS: Adults with diabetes are hospitalized patients with diabetes. Some readmissions, particularly and readmitted for a wide range of health conditions, for hypoglycemia and hyperglycemia, may be avoidable and hospitalizations for severe hypoglycemia and hyper- with better care transitions and post-discharge glycemia remain common, with high rates of recurrence. management. Severe dysglycemia is most likely to occur among younger OBJECTIVE: To ascertain the most common reasons and patients with multiple diabetes complications and prior risk factors for readmission among adults with diabetes, history of such events. with specific consideration of severe dysglycemia. KEY WORDS: DESIGN: Retrospective analysis of data from the Optum- diabetes; epidemiology; readmissions; hospital medicine; hypoglycemia; hyperglycemia; health services research; risk assessment. Labs Data Warehouse, an administrative data set of com- J Gen Intern Med 32(10):1097–105 mercially insured and Medicare Advantage beneficiaries DOI: 10.1007/s11606-017-4095-x across the U.S. © Society of General Internal Medicine 2017 PARTICIPANTS: Adults ≥18 years of age with diabetes, discharged from a hospital between January 1, 2009, and December 31, 2014 (N = 342,186). MAIN MEASURES: Principal diagnoses and risk factors INTRODUCTION for 30-day unplanned readmissions, subset as being for severe dysglycemia vs. all other causes. Diabetes increases the risk of hospitalization and unplanned KEY RESULTS: We analyzed 594,146 index hospitaliza- readmission,1–4 imposing a substantial burden on patients, care- tions among adults with diabetes: mean age 68.2 years givers, health systems, and the economy. Patients with diabetes (SD, 13.0), 52.9% female, and 67.8% white. The all-cause are susceptible to a broad range of acute illnesses that can lead 30-day readmission rate was 10.8%. Heart failure was the to hospitalization and readmission, including those related to most common cause for index hospitalization (5.5%) and readmission (8.9%). Severe dysglycemia accounted for co-existing chronic conditions as well as exacerbations of, and 2.6% of index hospitalizations (48.1% hyperglycemia, complications from, their underlying diabetes. Uniquely, 50.4% hypoglycemia, 1.5% unspecified) and 2.5% of read- patients with diabetes are vulnerable to severe hypoglycemia missions (38.3% hyperglycemia, 61.0% hypoglycemia, and hyperglycemia (Bsevere dysglycemia^), particularly during 0.7% unspecified). Younger patient age, severe dysglyce- the post-discharge period when changes in activity, nutrition, mia at index or prior hospitalization, and the Diabetes and medications commonly occur.5 Efforts to reduce hospital- Complications Severity Index (DCSI) were the strongest izations and readmissions among patients with diabetes require risk factors predisposing patients to severe dysglycemia vs. other readmissions. Prior episodes of severe dysglyce- a broad and nuanced understanding of their most common mia and the DCSI were also independent risk factors for causes, including the relative frequency of severe dysglycemic other-cause readmissions, irrespective of the cause of the events, as well as the risk factors for these events. index hospitalization. Prior studies of hospital readmissions among patients with diabetes are limited, but show that all-cause readmission risk is 1–3,6,7 Electronic supplementary material The online version of this article higher among patients with multiple comorbidities and (doi:10.1007/s11606-017-4095-x) contains supplementary material, among racial minorities.1–3,8,9 Several studies focused on hos- which is available to authorized users. pitalizations for hypoglycemia8,10 or hyperglycemia,8 but Received December 23, 2016 without placing them in the broader context of hospitalizations Revised April 19, 2017 Accepted May 31, 2017 and readmissions. Severe dysglycemic events can be avoided 11 Published online July 6, 2017 with optimal transitional and ambulatory care, and

1097 1098 McCoy et al.: Unplanned Readmissions among Adults with Diabetes JGIM understanding their relative importance may guide interven- hospitalization claim and grouped using categories identified tions for patients at high risk. Accordingly, our study addresses by the Agency for Healthcare Research and Quality (AHRQ) the following three questions in a large cohort of commercially Comorbidity and Clinical Classifications Software (CCS).18 insured and Medicare Advantage beneficiaries across the U.S.: Severe dysglycemic events were categorized as hypoglyce- (1) What are the principal reasons for which patients with mic, hyperglycemic, or unspecified (Online Appendix diabetes are hospitalized and subsequently readmitted? (2) Table 1).19–21 CCS 50 (Bdiabetes mellitus with complications^) What is the relative burden of hospitalizations and readmis- was not used, as it does not include hypoglycemia; it was sions for severe dysglycemia when considered in the context included in the analyses but modified to exclude diagnoses of all hospitalizations and readmissions experienced by this included in the severe dysglycemia definitions. patient population? (3) What are the primary risk factors for Patient demographic characteristics were obtained from unplanned 30-day readmissions in this high-risk population? enrollment data, and included age at the index hospitalization, This knowledge may facilitate the development and prioriti- sex, race/ethnicity, and annual household income in US dol- zation of individualized interventions for patients at highest lars. For each index hospitalization we determined the LOS risk for diabetes-related and unrelated readmissions. (categorized as 1, 2–4, 5–7, 8–14, and ≥15 days), discharge year, and history of another hospitalization during the prior 12 months. Rather than assume a linear or monotone relation- METHODS ship between LOS and risk of readmission, we categorized this risk factor based on inspection of the distribution for our Study Design and Data Source cohort. To determine whether a hospitalization was planned, We conducted a retrospective analysis of medical and phar- we applied version 3 of the CMS HWRM planned readmis- 16,22,23 macy claims data from the OptumLabs Data Warehouse sion algorithm. (OLDW), an administrative database of more than 100 million To identify comorbidities, we used the Diabetes Complica- 24,25 privately insured and Medicare Advantage enrollees through- tions Severity Index (DCSI), adapted for claims data, and 26 out the U.S. (Online Appendix 1).12,13 All study data were the Charlson Comorbidity Index (CCI), all from within accessed after the data had been de-identified, consistent with 12 months of the index admission. Treatment with insulin the Health Insurance Portability and Accountability Act was ascertained from pharmacy claims within 120 days pre- (HIPAA) expert de-identification determination and with ap- ceding the index admission. propriate legal, regulatory, and contractual permissions. Be- cause this study involved analysis of pre-existing, de- Outcomes identified data, it was exempt from institutional review board The primary outcome was unplanned readmission within approval.14 30 days of discharge for any cause, categorized as a readmission for severe dysglycemia vs. other (Online Appendix Table 1). We Study Population also conducted a secondary analysis which stratified patients The study sample comprised all index hospitalizations among based on their index hospitalization (severe dysglycemia vs. adults ≥18 years of age with diabetes in non-federal acute care other). Planned readmissions, defined using HWRM crite- hospitals lasting ≥1 night, with discharge alive date between ria,16,22,23 were excluded. January 1, 2009, and December 31, 2014 (Fig. 1). Diabetes was defined using Healthcare Effectiveness Data and Infor- Statistical Analysis mation Set (HEDIS) claims-computable criteria, excluding We summarized patient and hospitalization characteristics for gestational diabetes (ICD-9-CM 648.0, 648.8).15 Included each index admission, overall and by readmission status individuals had ≥12 months of enrollment before the index (none, severe dysglycemia, other), and tested for differences admission date and ≥31 days of enrollment after the index in outcomes using χ2 tests of independence. We identified the discharge date. In accordance with the Centers for Medicare & 10 most common principal diagnosis CCS categories18 at Medicaid Services (CMS) Hospital-Wide Readmission Mea- index admission, overall and by readmission status. sure (HWRM) and the National Committee for Quality As- In order to identify independent risk factors for severe surance criteria,16,17 hospitalizations for medical treatment of dysglycemia vs. other readmissions, we estimated a multino- cancer, psychiatric disease, and pregnancy were excluded. We mial logistic regression model using the primary outcome of excluded admissions with length of stay (LOS) >365 days and no readmission, other readmission, or severe dysglycemia readmissions within 1 day of discharge, which were counted readmission as the dependent variable. To account for hetero- as part of the original admission. geneity in index admission diagnoses between patients read- mitted for severe dysglycemia and for other reasons, and the Explanatory Variables contribution of these diagnoses to subsequent readmission Principal diagnoses for index admissions were identified using risk, we identified the 20 CCS groups that were most strongly the first/primary diagnosis ICD-9-CM code from the associated with readmission for any cause based on bivariate JGIM McCoy et al.: Unplanned Readmissions among Adults with Diabetes 1099

Figure 1 Study cohort. 1100 McCoy et al.: Unplanned Readmissions among Adults with Diabetes JGIM analyses for all 46 CCS groups with a prevalence of at least hospitalizations are presented in Figure 2, highlighting the 0.5% in the study cohort, and included these as risk factors in frequency of readmissions for heart failure, sepsis, and acute all models (Online Appendix Table 2). These 20 CCS groups renal failure in both cohorts. were included in the multinomial logistic regression model, together with the explanatory variables listed above. We used the Huber–White sandwich variance estimator to account for General Risk Factors for Readmission correlation of outcomes by hospital.27,28 To account for miss- Certain variables increased readmission risk globally, for both ing income data (8.1% of total sample), we used multiple 29 severe dysglycemia and other-cause readmissions. Longer stochastic imputations, with 20 imputations. We report mod- LOS during the index hospitalization increased the risk of all el effects as relative risk ratios (RRRs). To better interpret readmissions. Readmission risk was significantly lower fol- differences in RRRs across severe dysglycemia vs. other read- lowing planned rather than unplanned hospitalizations: RRR missions, we calculated ratios of RRRs for severe dysglycemia 0.57 (95% CI, 0.47–0.69) for severe dysglycemia readmission vs. other-cause readmissions, and the p-values for the test in and 0.79 (95% CI, 0.76–0.81) for other readmissions. Higher which the ratio of RRRs differed from 1.0; that is, that the income was associated with reduced risk of all readmissions, RRR for severe dysglycemia readmissions was not the same as with no significant difference in effect on the two types of the RRR for other readmissions. This P-value was based on a readmission. African American patients had a slightly higher test which incorporated the variances in the RRRs. Thus, a risk of all readmissions than did white patients. severe dysglycemia/other RRR ratio >1 indicates that the particular covariate increases severe dysglycemia readmission Readmissions for Severe Dysglycemia risk more than other-cause readmission risk, and vice versa. All analyses were conducted using Stata 14.1 (2015; Stata- Many of the examined covariates had different effects on severe Corp LP, College Station, TX). P-values <0.05 were consid- dysglycemia vs. other-cause readmissions (Table 2). The stron- ered statistically significant. gest risk factors for severe dysglycemia readmissions were patient age, DCSI, index hospitalization for severe dysglycemia (hyperglycemic events had a greater impact than hypoglycemic RESULTS events), and prior history of severe dysglycemia hospitaliza- tions (Table 2). The risk of readmission for severe dysglycemia Study Population and 30-Day Unplanned was nearly twice as high among patients aged <45 years as Readmission those ≥45 years; age was not a strong risk factor for other-cause Between January 1, 2009, and December 31, 2014, 342,186 readmissions until age ≥75 years, when it began to rise. For all adults with diabetes experienced 594,146 index hospitaliza- age groups, age was a stronger determinant of severe dysgly- tions, including 15,644 hospitalizations for severe dysglyce- cemia readmissions than other-cause readmissions. The DCSI mia (2.6% of index hospitalizations) among 13,161 patients was a significant risk factor for both categories of readmissions, (Table 1). Of these, 48.1% were for hypoglycemia, 50.4% for but with greater impact on severe dysglycemia. Compared to hyperglycemia, and 1.5% were unspecified. The most com- patients with no diabetes complications, the risk of readmission mon cause for index hospitalization was heart failure (5.5% of for severe dysglycemia was 2.33-fold higher among patients index hospitalizations), while severe dysglycemic events were with DCSI of 3–6, and 3.20-fold higher with DCSI ≥7. Com- the 11th most common cause; the top 20 of 2621 diagnosis pared to non-insulin-treated patients, patients using insulin were categories at index hospitalization are shown in Online Ap- 80% more likely to be readmitted for severe dysglycemia and pendix Table 3. 6% more likely to be readmitted for other causes. The all-cause unplanned readmission rate was 10.8%. Heart Patients whose index hospitalization was for severe dysgly- failure was the most common reason for readmission (8.9% of cemia were nearly nine times as likely to be readmitted for readmissions), while severe dysglycemic events accounted for recurrent severe dysglycemia than for another reasons. Among 2.5% of readmissions (38.3% hyperglycemia, 61.0% hypogly- patients hospitalized for hypoglycemia, the RRR for severe cemia, 0.7% unspecified). The reason for readmission varied dysglycemia readmission was 4.74 (95% CI, 3.81–5.89) and based on whether the index hospitalization was for severe other-cause readmission was 0.87 (95% CI, 0.81–0.94), with a dysglycemia or for other causes (Fig. 2; Online Appendix severe dysglycemia/other RRR ratio of 5.43. Severe hypergly- Tables 4, 5). Of those hospitalized for severe hypoglycemia, cemia index hospitalizations had an even greater impact on 6.7% were readmitted for another severe dysglycemic event severe dysglycemia readmission risk, with an RRR of severe and 1.1% were readmitted for other causes. Of those hospital- dysglycemia readmission of 8.57 (95% CI, 7.08–10.37) and ized for severe hyperglycemia, 20.8% were readmitted for other-cause readmission of 0.60 (95% CI, 0.55–0.67); the severe another severe dysglycemic event and 0.7% were readmitted dysglycemia/other RRR ratio was 15.89. Having prior hospital- for other causes. Only 1.8% of patients hospitalized for other izations for severe dysglycemia was also a risk factor for all causes were readmitted for severe dysglycemia. Other causes readmissions, but readmission for severe dysglycemia was near- for readmission following severe dysglycemia and other-cause ly four times as likely: RRR for severe dysglycemia readmission JGIM McCoy et al.: Unplanned Readmissions among Adults with Diabetes 1101

Table 1 Study Population

Study cohort Not readmitted Readmitted P-value N(%) N(%) Severe dysglycemia Other cause N(%) N(%)

Index discharges 594,146 524,190 1744 68,212 Patient characteristics* Age, mean (SD) 68.2 (13.0) 68.0 (13.0) 63.9 (18.0) 69.7 (12.2) Age, N (%) <0.001 <45 30,370 (5.1) 27,694 (5.3) 279 (16.0) 2397 (3.5) 45–64 179,924 (30.3) 161,069 (30.7) 452 (25.9) 18,403 (27.0) 65–74 173,279 (29.2) 152,492 (29.1) 431 (24.7) 20,356 (29.8) ≥75 210,573 (35.4) 182,935 (34.9) 582 (33.4) 27,056 (39.7) Sex 0.83 Male 279,651 (47.1) 246,787 (47.1) 812 (46.6) 32,052 (47.0) Female 314,495 (52.9) 277,403 (52.9) 932 (53.4) 36,160 (53.0) Race <0.001 White 402,747 (67.8) 356,824 (68.1) 1090 (62.5) 44,833 (65.7) Black 109,670 (18.5) 95,059 (18.1) 445 (25.5) 14,166 (20.8) Asian 12,344 (2.1) 10,956 (2.1) 38 (2.2) 1350 (2.0) Hispanic 47,471 (8.0) 41,910 (8.0) 133 (7.6) 5428 (8.0) Other/unknown 21,914 (3.7) 19,441 (3.7) 38 (2.2) 2435 (3.6) Annual income (US$) <0.001 <40,000 242,291 (40.8) 210,857 (40.2) 795 (45.6) 30,639 (44.9) 40,000–49,000 58,233 (9.8) 51,567 (9.8) 152 (8.7) 6514 (9.5) 50,000–59,000 49,511 (8.3) 43,775 (8.4) 142 (8.1) 5594 (8.2) 60,000–74,000 55,557 (9.4) 49,492 (9.4) 158 (9.1) 5907 (8.7) 75,000–99,000 61,542 (10.4) 55,114 (10.5) 159 (9.1) 6269 (9.2) ≥100,000 78,364 (13.2) 71,033 (13.6) 180 (10.3) 7151 (10.5) Missing 48,648 (8.2) 42,352 (8.1) 158 (9.1) 6138 (9.0) Index hospitalization characteristics Planned status <0.001 Unplanned 500,750 (84.3) 438,277 (83.6) 1613 (92.5) 60,860 (89.2) Planned 93,396 (15.7) 85,913 (16.4) 131 (7.5) 7352 (10.8) Reason for index hospitalization Severe dysglycemia 15,644 (2.6) 13,970 (2.7) 483 (27.7) 1191 (1.7) Hypoglycemia 7528 (1.3) 6686 (1.3) 116 (6.7) 726 (1.1) <0.001 Hyperglycemia 7886 (1.3) 7071 (1.3) 363 (20.8) 452 (0.7) <0.001 Unspecified 230 (0.0) 213 (0.0) * (*) 13 (0.0) <0.001 Other causes 578,502 (97.4) 510,220 (97.3) 1261 (72.3) 67,021 (98.3) Length of stay <0.001 1 day 80,388 (13.5) 74,637 (14.2) 169 (9.7) 5582 (8.2) 2–4 days 314,962 (53.0) 283,483 (54.1) 822 (47.1) 30,657 (44.9) 5–7 days 118,193 (19.9) 101,330 (19.3) 403 (23.1) 16,460 (24.1) 8–14 days 63,392 (10.7) 51,613 (9.8) 265 (15.2) 11,514 (16.9) ≥15 days 17,211 (2.9) 13,127 (2.5) 85 (4.9) 3999 (5.9) Prior hospitalizations None <0.001 Other causes only 351,468 (59.2) 321,860 (61.4) 693 (39.7) 28,915 (42.4) Severe dysglycemia 236,144 (39.7) 196,958 (37.6) 738 (42.3) 38,448 (56.4) Diabetes characteristics Diabetes Complications Severity Index <0.001 0 143,755 (24.2) 133,250 (25.4) 153 (8.8) 10,352 (15.2) 1–2 152,519 (25.7) 137,598 (26.2) 333 (19.1) 14,588 (21.4) 3–6 198,156 (33.4) 172,002 (32.8) 695 (39.9) 25,459 (37.3) 7–29 99,716 (16.8) 81,340 (15.5) 563 (32.3) 17,813 (26.1) Diabetes treatment <0.001 Non-insulin only 435,861 (73.4) 387,501 (73.9) 47,563 (69.7) 797 (45.7) Insulin 158,285 (26.6) 136,689 (26.1) 20,649 (30.3) 947 (54.3) *Baseline patient characteristics at the time of index hospitalization. Each index admission was considered separately, and data were ascertained at the time of discharge of 5.47 (95% CI, 4.60–6.51) and other-cause readmission of remained unchanged, with few notable differences. While 1.43 (95% CI, 1.33–1.55) compared to no prior hospitalizations. higher DCSI increased the risk of all readmissions following all index hospitalizations, it had the greatest impact on read- mission risk following index stays for other (not severe dys- Readmission Following Severe Dysglycemic vs. glycemia) causes: DCSI ≥7 (compared to DCSI = 0) increased Other Events the risk of severe dysglycemia readmissions by RRR 3.66 In a secondary analysis examining patients hospitalized for (95% CI, 2.87–4.66) and other-cause readmissions by RRR severe dysglycemia separately from those hospitalized for 1.77 (95% CI, 1.71–1.82) after an index hospitalization for other causes (Online Appendix Tables 6 and 7), most risk other causes, compared to RRR 1.91 (95% CI, 1.27–2.88) and factors for severe dysglycemia and other-cause readmissions RRR 2.01 (95% CI, 1.58–2.56) for severe dysglycemia and 1102 McCoy et al.: Unplanned Readmissions among Adults with Diabetes JGIM

Figure 2 Principal diagnoses at readmission. We present the percentage of patients readmitted for a specific principal diagnosis CCS category among all patients who were hospitalized for either a severe dysglycemic event (dark gray) or another cause (light gray) as their index hospitalization. The 20 most prevalent readmission diagnosis categories overall are presented. DM diabetes mellitus, UTI urinary tract infection, GI gastrointestinal, MI myocardial infarction, COPD chronic obstructive pulmonary disease. other-cause readmissions, respectively, following index hos- cause for readmission following non-diabetes-related index hos- pitalization for severe dysglycemia. Longer LOS did not affect pitalizations, and the most common cause for readmission fol- the risk of readmission for severe dysglycemia following lowing an index hospitalization for severe dysglycemia. index severe dysglycemic events, while increasing it follow- Younger patients were the most likely to be readmitted for ing index hospitalizations for other causes. severe dysglycemia, with 10% of all readmissions among patients aged 18–44 years being for severe dysglycemia vs. just over 2% in the older age groups. This may reflect greater clinical complexity and multi-morbidity of older patients, placing them at DISCUSSION risk for multiple diseases and complications unrelated to diabetes. The all-cause readmission rate among 342,186 U.S. adults with Younger patients may have worse glycemic control or greater diabetes was 10.8%, reflective of the wide age and morbidity glycemic variability, both of which raise the risk of severe distribution of this commercially and Medicare Advantage- dysglycemia but cannot be captured by administrative data. insured population. The vast majority, 97.4%, of the index Younger patients are also more likely to have type 1 diabetes, hospitalizations were for causes not directly related to diabetes, though administrative data cannot reliably identify diabetes type. with only 2.6% being primarily for the management of severe Conversely, older patients may be treated less intensively and/or dysglycemia (48.1% hypoglycemia and 50.4% hyperglycemia). have greater attention paid to their diabetes management during Similarly, 97.5% of all readmissions were for causes unrelated to and following hospitalization, which would lower their risk of short-term glycemic control, and 2.5% were specifically for severe dysglycemia. Similarly, older patients with complex health severe dysglycemia, though the relative frequency of hypogly- or social needs may have closer outpatient follow-up after dis- cemic and hyperglycemic events was reversed, with 61.0% of charge, decreasing their risk of unnoticed and unaddressed gly- these events for hypoglycemia and 38.3% for hyperglycemia. cemic deterioration. These variables could not be assessed in the The most common cause for hospitalization and readmission context of this study, but would benefit from further examination. among these diabetes patients was heart failure, which is consis- An index or prior hospitalization for severe dysglycemia was tent with studies conducted in the general U.S. population.4 Still, the strongest predictor of readmission for severe dysglycemia. despite being potentially avoidable with optimal ambulatory and Prior hospitalizations for severe dysglycemia increased the risk transitional care,11 severe dysglycemia was the 11th most com- of readmission for severe dysglycemia sixfold, irrespective of the mon cause for index hospitalization, the 14th most common reason for the index hospitalization. Additionally, nearly one- JGIM McCoy et al.: Unplanned Readmissions among Adults with Diabetes 1103

Table 2 Risk Factors for Readmission for Severe Dysglycemia and Other Causes

Readmission for severe Readmission for other causes Wald P* Ratio of severe dysglycemia dysglycemia/other† RRR (95% CI) P RRR (95% CI) P P

Patient characteristics Age, years <0.001 <45 Ref Ref Ref 45–64 0.52 (0.43, 0.64) <0.001 0.98 (0.94, 1.03) 0.42 0.54 <0.001 65–74 0.56 (0.46, 0.69) <0.001 1.00 (0.95, 1.06) 0.96 0.56 <0.001 ≥75 0.60 (0.49, 0.74) <0.001 1.06 (1.00, 1.12) 0.04 0.57 <0.001 Sex 0.93 Male Ref Ref Ref Female 0.98 (0.88, 1.09) 0.701 1.00 (0.98, 1.02) 0.99 0.98 0.70 Race <0.001 White Ref Ref Ref Black 1.22 (1.07, 1.38) 0.002 1.09 (1.06, 1.12) <0.001 1.12 0.10 Asian 1.30 (0.95, 1.78) 0.106 1.05 (0.98, 1.11) 0.15 1.24 0.23 Hispanic 0.98 (0.81, 1.18) 0.817 1.05 (1.01, 1.09) 0.02 0.94 0.46 Other/unknown 0.69 (0.50, 0.97) 0.03 1.01 (0.96, 1.06) 0.79 0.69 0.008 Annual income (US$) <0.001 <40,000 Ref Ref Ref 40–49,000 0.77 (0.64, 0.93) 0.007 0.94 (0.91, 0.97) <0.001 0.82 0.03 50–59,000 0.94 (0.79, 1.12) 0.49 0.98 (0.95, 1.01) 0.27 0.96 0.62 60–74,000 0.89 (0.74, 1.07) 0.21 0.95 (0.92, 0.98) 0.001 0.94 0.47 75–99,000 0.79 (0.66, 0.95) 0.01 0.94 (0.91, 0.97) <0.001 0.84 0.04 ≥100,000 0.77 (0.65, 0.91) 0.003 0.90 (0.87, 0.93) <0.001 0.85 0.05 Index hospitalization characteristics Reason for index hospitalization <0.001 Other causes Ref Ref Ref Hypoglycemia 4.74 (3.81, 5.89) <0.001 0.87 (0.81, 0.94) 0.001 5.434 <0.001 Hyperglycemia 8.57 (7.08, 10.37) <0.001 0.60 (0.55, 0.67) <0.001 14.188 <0.001 Unspecified 5.81 (2.00, 16.83) 0.001 0.61 (0.35, 1.07) 0.082 9.564 0.137 Planned status <0.001 Unplanned Ref Ref Ref Planned 0.57 (0.47, 0.69) <0.001 0.79 (0.76, 0.81) <0.001 0.73 <0.001 Length of stay <0.001 1 day 0.79 (0.67, 0.93) 0.006 0.77 (0.75, 0.79) <0.001 1.03 0.77 2–4days Ref Ref Ref 5–7 days 1.42 (1.25, 1.62) <0.001 1.35 (1.32, 1.38) <0.001 1.06 0.43 8–14 days 1.85 (1.60, 2.15) <0.001 1.79 (1.74, 1.83) <0.001 1.04 0.64 ≥15 days 2.42 (1.91, 3.07) <0.001 2.45 (2.35, 2.55) <0.001 0.99 0.93 Prior hospitalizations <0.001 None Ref Ref Ref Other causes 1.27 (1.13, 1.42) <0.001 1.58 (1.55, 1.61) <0.001 0.81 <0.001 Severe dysglycemia 5.47 (4.60, 6.51) <0.001 1.43 (1.33, 1.55) <0.001 3.82 <0.001 Diabetes characteristics Diabetes Complications Severity Index <0.001 0Ref Ref Ref 1–2 1.52 (1.25, 1.85) <0.001 1.19 (1.15, 1.22) <0.001 1.28 0.03 3–6 2.33 (1.94, 2.81) <0.001 1.41 (1.37, 1.45) <0.001 1.65 <0.001 7–29 3.20 (2.60, 3.94) <0.001 1.77 (1.72, 1.83) <0.001 1.81 <0.001 Diabetes treatment <0.001 Non-insulin Ref Ref Ref Insulin 1.79 (1.61, 1.99) <0.001 1.06 (1.04, 1.08) <0.001 1.68 <0.001 Note: Adjusted multinomial logit regression model showing the relative risk ratios (RRRs) for severe dysglycemia and other-cause 30-day readmission compared to no readmission. The severe dysglycemia/other ratio is a ratio of the RRRs of severe dysglycemia and other-cause readmission. Specifically, a severe dysglycemia/other ratio >1 signifies a greater impact of the covariate on readmissions for severe dysglycemia than for other causes, and vice versa. The model was adjusted for the Charlson comorbidities (dementia, chronic obstructive pulmonary disease, rheumatologic disease, peptic ulcer disease, liver disease, hemiplegia/paraplegia, cancer, and AIDS/HIV), the discharge year, and the 20 principal diagnosis CCS categories from the index discharge with the highest risk of all-cause 30-day readmission *The Wald p-value is a test of overall (joint) effect of multi-category variables †P-value for a two-sided test of the severe dysglycemia/other RRR being different from 1

third of patients whose index hospitalization was for severe dysglycemia readmissions, these hospitalizations also increased dysglycemia were later readmitted for a recurrent severe dysgly- the risk of other-cause readmissions, suggesting that severe cemic event. The risk of recurrent dysglycemia was higher after dysglycemia may be a global marker of patient complexity and an index hyperglycemic event (8.6-fold increase compared to illness. The DCSI and insulin therapy had similar effects on index hospitalizations for other causes) than hypoglycemic event severe dysglycemia and other-cause readmissions, with readmis- (4.7-fold increase). Still, while prior hospital stays for severe sion risk increasing linearly with the DCSI but more steeply for dysglycemia had a very strong impact on the likelihood of severe dysglycemia-specific readmissions. This reinforces the fact that 1104 McCoy et al.: Unplanned Readmissions among Adults with Diabetes JGIM diabetes severity is an important consideration for all readmis- and those with diabetes complications are at highest risk for sions, not just those directly related to glycemic control. severe dysglycemia readmission, and may benefit from Hospitalization is likely to affect the post-discharge diabetes diabetes-specific interventions. History of severe dysglycemia management and overall health of patients with diabetes, and and burden of diabetes complications are also independent risk thereby increases the risk for readmission for severe dysglyce- factors for all-cause readmission, and should be considered in mia, irrespective of the reason for the index admission. Diabetes the management of all hospitalized patients with diabetes. management is affected by the patient’s nutrition, physical ac- tivity, sleep, pain, and non-diabetic drugs—all of which are Corresponding Author: Rozalina G. McCoy, MD MS; Division of potentially altered during and after hospitalization. This, in turn, Primary Care Internal Medicine, Department of MedicineMayo Clinic, Rochester, MN 55905, USA (e-mail: [email protected]). may result in increased susceptibility to severe dysglycemic events. While we initially hypothesized that the risk of readmis- Author Contributions McCoy and Shah had full access to all the sion for severe dysglycemia might be elevated following index data in the study and take responsibility for the integrity of the data hospitalization for other causes due to the relative de- and the accuracy of the data analysis. Study conception and design: prioritization of diabetes management in the context of an unre- McCoy, Shah. Acquisition of data: Herrin, Jeffrey, Shah. Interpretation of data: McCoy, Lipska, Herrin, Shah. Drafting of the manuscript: lated acute illness, this was not substantiated by the data. Patients McCoy. Critical revision of the manuscript for important intellectual hospitalized for causes other than severe dysglycemia were less content: McCoy, Lipska, Herrin, Krumholz, Jeffrey, Shah. Statistical likely to be readmitted for severe dysglycemia, while those analysis: Herrin. Obtaining funding: Shah. Administrative, technical, and material support: N/A. Study supervision: Shah. hospitalized for severe dysglycemia had the greatest risk of readmission for recurrent severe dysglycemia. This observation Compliance with Ethical Standards: is important, because it reinforces both the relative frequency and Funders: This publication was made possible by CTSA Grant Number the high probability of recurrent severe dysglycemia, despite that UL1 TR000135 from the National Center for Advancing Translational fact that these events may be avoidable. Inpatient diabetes edu- Sciences (NCATS), a component of the National Institutes of Health 30 (NIH). Its contents are solely the responsibility of the authors and do cation has been shown to reduce the risk of readmission and not necessarily represent the official views of the NIH. Dr. Shah was may benefit high-risk patients. Similarly, medication review and funded in part through project number R18HS18339 from the Agency for Healthcare Research and Quality (AHRQ). Dr. McCoy receives sup- reconciliation, care transition programs, and follow-up telephone port from the Mayo Clinic Robert D. and Patricia E. Kern Center for the calls or appointments that specifically address diabetes manage- Science of Health Care Delivery. Dr. Lipska receives support from the ment may reduce the risk of readmission for severe dysglycemia. National Institute on Aging and the American Federation of Aging Research through the Paul Beeson Career Development Award Such interventions could not be identified in our data set, and (K23AG048359) and the Yale Claude D. Pepper Older Americans Inde- accounting for these programs or any outpatient follow-up after pendence Center (P30AG021342). She also receives support from CMS discharge would be problematic due to endogeneity. to develop and maintain publicly reported quality measures. This study needs to be considered in light of its limitations Prior Presentations: Parts of this work were previously presented as stemming from the use of administrative data, which cannot poster presentations at the American Diabetes Association 75th Scien- tific Sessions in Boston, MA (June 7, 2015), and the AcademyHealth capture all potential risk factors for severe dysglycemia and Annual Research Meeting in Minneapolis, MN (June 15, 2015). readmission, including clinical information (e.g., hemoglobin A ),31 frailty,32 social support, and socioeconomic factors. We Conflict of Interest: Dr. Krumholz is a recipient of research agree- 1c ments from Medtronic and Johnson & Johnson (Janssen), through Yale, classified hospitalizations on the basis of principal discharge to develop methods of clinical trial data sharing; is the recipient of a diagnoses, which may underestimate the incidence of severe grant from Medtronic and the Food and Drug Administration, through Yale, to develop methods for post-market surveillance of medical devi- dysglycemia by capturing only primary dysglycemic events rath- ces; works under contract with the Centers for Medicare & Medicaid er than severe dysglycemia precipitated by another potentially Services to develop and maintain performance measures; chairs a unrelated illness. Because patients were required to have ≥31 days cardiac scientific advisory board for UnitedHealth; is a participant/ participant representative of the IBM Watson Health Life Sciences of enrollment following index hospitalization, patients who died Board; is a member of the Advisory Board for Element Science, the during this period were excluded, even though their death may Physician Advisory Board for Aetna, and the Open Trials Advisory Board for the Laura and John Arnold Foundation; and is the founder have been informative. Our study was also limited to people with of Hugo, a personal health information platform. All remaining authors commercial or Medicare Advantage health plans, and did not declare that they do not have a conflict of interest. include patients who were uninsured or had public health cover- age. The population studied here fills the gap in readmissions literature that has heretofore focused predominantly on selected states or health systems,1–4,33 or on the Medicare fee-for-service population,6,8,9 as it includes people from a wide range of ages, REFERENCES racial/ethnic groups, health systems, and geographic regions. 1. Jiang HJ, Stryer D, Friedman B, Andrews R. Multiple hospitalizations These findings constitute novel and important contributions for patients with diabetes. Diabetes Care. 2003;26(5):1421–1426. 2. Jiang HJ, Andrews R, Stryer D, Friedman B. Racial/ethnic disparities to the management of diabetes. While patients with diabetes in potentially preventable readmissions: the case of diabetes. Am J Public are hospitalized and readmitted for a large number of health Health. 2005;95(9):1561–1567. 3. Kim H, Ross JS, Melkus GD, Zhao Z, Boockvar K. Scheduled and conditions, hospitalizations for severe dysglycemia remain unscheduled hospital readmissions among patients with diabetes. Am J common and have a high rate of recurrence. Younger patients Manag Care. 2010;16(10):760–767. JGIM McCoy et al.: Unplanned Readmissions among Adults with Diabetes 1105

4. Donze J, Lipsitz S, Bates DW, Schnipper JL. Causes and patterns of for a principal diagnosis of diabetes with short-term complications per readmissions in patients with common comorbidities: retrospective 100,000 population, ages 18 years and older. In: National Quality cohort study. BMJ. 2013;347:f7171. Measures Clearinghouse (NQMC) [Web site]. Rockville (MD): Agency for 5. Krumholz HM. Post-hospital syndrome–an acquired, transient condition Healthcare Research and Quality (AHRQ); March 1, 2015. Available at: of generalized risk. N Engl J Med. 2013;368(2):100–102. https://www.qualitymeasures.ahrq.gov/summaries/summary/49498. 6. Raval AD, Zhou S, Wei W, Bhattacharjee S, Miao R, Sambamoorthi U. Accessed on May 18, 2017. 30-Day Readmission Among Elderly Medicare Beneficiaries with Type 2 20. National Quality Measures Clearinghouse (NQMC). Measure summary: Diabetes. Popul Health Manag. 2015. Uncontrolled diabetes admission: percentage of admissions for a princi- 7. Eby E, Hardwick C, Yu M, et al. Predictors of 30 day hospital pal diagnosis of diabetes without mention of short-term or long-term readmission in patients with type 2 diabetes: a retrospective, case- complications per 100,000 population, ages 18 years and older. In: control, database study. Curr Med Res Opin. 2015;31(1):107–114. National Quality Measures Clearinghouse (NQMC) [Web site]. Rockville 8. Lipska KJ, Ross JS, Wang Y, et al. National trends in US hospital (MD): Agency for Healthcare Research and Quality (AHRQ); March 1, admissions for hyperglycemia and hypoglycemia among Medicare bene- 2015. Available at: https://www.qualitymeasures.ahrq.gov/summaries/ ficiaries, 1999 to 2011. JAMA Intern Med. 2014;174(7):1116–1124. summary/49503. Accessed on May 18, 2017. 9. O’Neil SS, Lake T, Merrill A, Wilson A, Mann DA, Bartnyska LM. Racial 21. Ginde AA, Blanc PG, Lieberman RM, Camargo CA, Jr. Validation of disparities in hospitalizations for ambulatory care-sensitive conditions. ICD-9-CM coding algorithm for improved identification of hypoglycemia Am J Prev Med. 2010;38(4):381–388. visits. BMC Endocr Disord. 2008;8:4. 10. Emons MF, Bae JP, Hoogwerf BJ, Kindermann SL, Taylor RJ, 22. 2015 Measure Updates and Specifications Report Hospital-Wide All- Nathanson BH. Risk factors for 30-day readmission following Cause Unplanned Readmission Measure – Version 4.0. Submitted by Yale hypoglycemia-related emergency room and inpatient admissions. BMJ New Haven Health Services Corporation/Center for Outcomes Research Open Diabetes Res Care. 2016;4(1):e000160. & Evaluation (YNHHSC/CORE). Prepared for Centers for Medicare & 11. AHRQ QITM Research Version 5.0, Prevention Quality Indicators #1, Medicaid Services (CMS). March 2015. 22.; http://www.cms.gov/Medi- Technical Specifications, Diabetes Short-Term Complications Admission care/Quality-Initiatives-Patient-Assessment-Instruments/HospitalQuali- Rate. March 2015. Available at: http://www.qualityindicators.ahrq.gov/. tyInits/Measure-Methodology.html. Accessed May 18, 2017. Accessed May 18, 2017. 23. CMS. Centers for Medicare & Medicaid Services. Hospital Quality 12. Optum. Optum Research Data Assets. June 2015. Available at: https:// Initiative. Measure Methodology. 2016; https://www.cms.gov/Medi- www.optum.com/content/dam/optum/resources/productSheets/ care/Quality-Initiatives-Patient-Assessment-Instruments/HospitalQuali- 5302_Data_Assets_Chart_Sheet_ISPOR.pdf. Accessed May 18, 2017. tyInits/Measure-Methodology.html. Accessed May 18, 2017. 13. Wallace PJ, Shah ND, Dennen T, Bleicher PA, Crown WH. Optum Labs: 24. Young BA, Lin E, Von Korff M, et al. Diabetes complications severity Building A Novel Node In The Learning Health Care System. Health Aff index and risk of mortality, hospitalization, and healthcare utilization. (Millwood). 2014;33(7):1187–1194. Am J Manag Care. 2008;14(1):15–23. 14. Privacy of Individually Identifiable Health Information: Other require- 25. Chang HY, Weiner JP, Richards TM, Bleich SN, Segal JB. Validating ments relating to uses and disclosures of protected health information. the adapted Diabetes Complications Severity Index in claims data. Am J 45 CFR 164.514(b)(1) (June 7, 2013). Available at: http://www.ecfr.gov/ Manag Care. 2012;18(11):721–726. cgi-bin/retrieveECFR?gp=1&SID=ff85fb032e00c1b6d135ecf0767- 26. Charlson ME, Pompei P, Ales KA, MacKenzie CR. A new method of f18a0&ty=HTML&h=L&mc=true&r=PART&n=pt45.1.164#se45.1.164_ classifying prognostic comorbidity in longitudinal studies: Development 1514. Accessed May 18, 2017. and validation. Journal of Chronic Diseases. 1987;40(5):373–383. 15. National Committee for Quality Assurance (NCQA) Healthcare Effectiveness 27. White HLJ. A heteroskedasticity-consistent covariance matrix estimator Data and Information Set (HEDIS) Comprehensive Diabetes Care. Wash- and a direct test for heteroskedasticity. Econometrica. 1980;48:817–838. ington, D.C.: National Committee for Quality Assurance; 2015:74–98. 28. Huber PJ. The behavior of maximum likelihood estimates under 16. Horwitz LI, Grady JN, Dorsey KB, et al. 2014 Measures Updates and nonstandard conditions. Vol 1. Berkeley: University of California Press; Specifications Report: Hospital-Wide All-Cause Unplanned Readmission 1967. (Version 3.0). Submitted by Yale New Haven Health Services Corporation/ 29. Rubin DB. Multiple Imputation for Nonresponse in Surveys. New York: Center for Outcomes Research & Evaluation (YNHHSC/CORE). Prepared Wiley; 1987. for Centers for Medicare & Medicaid Services (CMS). Available at file:/// 30. HealySJ,BlackD,HarrisC,LorenzA,DunganKM.Inpatient diabetes C:/Users/m070233/Downloads/Hosp-Wide_Rdmsn_Msr_Updts_ education is associated with less frequent hospital readmission among 032114.pdf2014. patients with poor glycemic control. Diabetes Care. 2013;36(10):2960– 17. National Quality Measures Clearinghouse. All-cause readmissions: the 2967. number of acute inpatient stays during the measurement year that were 31. Tomlin AM, Dovey SM, Tilyard MW. Risk factors for hospitalization followed by an acute readmission for any diagnosis within 30 days and due to diabetes complications. Diabetes Res Clin Pract. 2008;80(2):244– the predicted probability of an acute readmission, for patients 18 years of 252. age and older. 2015; http://www.qualitymeasures.ahrq.gov/content. 32. Morandi A, Bellelli G, Vasilevskis EE, et al. Predictors of rehospitaliza- aspx?id=48857. Accessed May 18, 2017. tion among elderly patients admitted to a rehabilitation hospital: the role 18. HCUP Clinical Classifications Software (CCS) for ICD-9-CM. Healthcare of polypharmacy, functional status, and length of stay. J Am Med Dir Cost and Utilization Project (HCUP). Rockville: Agency for Healthcare Assoc. 2013;14(10):761–767. Research and Quality; 2006–2009. 33. Bennett KJ, Probst JC, Vyavaharkar M, Glover SH. Lower rehospital- 19. National Quality Measures Clearinghouse (NQMC). Measure summary: ization rates among rural Medicare beneficiaries with diabetes. J Rural Diabetes short term complications admission: percentage of admissions Health. 2012;28(3):227–234. End-of-Life Care Conversations: Medicare Reimbursement FAQs

The changes in Medicare reimbursement policy that went into effect January 2016 provide an opportunity for more clinicians and patients to engage in conversations about preferences for care at the end of life. However, many people are confused about where to start. Whether you are uncertain about the new rules for CMS reimbursements or about starting those conversations with patients, this document will help you understand this new landscape for end-of-life care conversations. Before getting started, check to see if a local coverage determination has been made, and check with your local billing expert to ensure your practice is compliant with their recommendations. Make sure that the new reimbursement codes have been added to your system’s billing apparatus. These codes may not be available until your facility approves them for use.

1. Do these new codes need to be used in the 7. How much do payers reimburse for these codes? context of an illness? 99497 = 1.5 RVUs No. In fact, any medical management must 99498 = 1.4 RVUs be billed separately. 8. Are there limits to the number of times that the 2. What are the new advance care planning (ACP) new codes can be used? codes from CMS that became active in 2016? There are no limits to the number of times the codes 99497 – ACP, including the explanation and discussion can be used. ACP can be readdressed as needed with a of advance directives, such as standard forms (with change in condition. Each time they are used, 99497 completion of such forms, when performed), by the should be used for the first 30 minutes and 99498 physician or other qualified health professional should be used for each additional 30 minutes. 99498 – Each additional 30 minutes (list separately in 9. Which health care providers can be reimbursed addition to code for primary procedure) for having ACP discussions with patients under the new rule? Can physicians charge for the 3. How much time must be spent to use the new codes? codes if another staff member engages the More than half of each interval must be used. For example: patient in the ACP discussion? • Use 99497 if you meet or exceed 16 minutes. Physicians (MDs and DOs), nurse practitioners (NPs), • Use 99497 + 99498 if you meet or exceed 46 minutes. and physician assistants (PAs) (i.e. those who are authorized to independently bill Medicare for Current 4. Does the conversation have to be in-person to use Procedural Terminology (CPT) services) are the only the new codes? Does it have to be with the patient? providers who can use these codes. The conversation has to be in-person (you cannot use the “Incident to” rules apply in the outpatient setting. code for telehealth), but it doesn’t have to be with the This means that a provider can use these codes patient. It can be with a surrogate or family members. if they perform an initial service and a non-billing 5. What are the documentation requirements? team member (e.g., registered nurse, social worker) helps deliver part of the service, with ongoing direct • Total time in minutes supervision and involvement of the billing provider. • Patient/surrogate/family “given opportunity to decline” Example: The physician starts an ACP conversation, • Details of content (e.g. Who was involved? What was then says, “I’d like to introduce you to our nurse who discussed? Understanding of illness, spiritual factors. will talk with you about choosing a surrogate medical Why are they making the decisions they are making? decision maker and discuss with you how you might Was any advance directive offered/filled out? Follow-up) have a conversation with that person,” then debriefs 6. What costs might patients incur from these codes? afterwards with the patient. Work with your local billing expert regarding “incident to” rules. When a provider discusses advance care planning with 10. How can physicians bill for these conversations a patient at his/her Annual Wellness Visit, there is no for non-Medicare patients? cost to the patient. However, if the provider has an ACP conversation at other times, Part B cost sharing applies If the patient has private insurance, find out if ACP and the patient may be responsible for copay/coinsurance. conversations are covered. Otherwise, you can use “counseling and coordination of care” codes, but only in the context of a serious illness. Having the Conversation at Three Life Stages: A Guide for Providers

No Serious Illness Serious Illness Advanced Serious Illness Ms. Smith is a 68-year-old woman At age 71, Ms. Smith developed a Now 75 years old, Ms. Smith has with hypertension, hyperlipidemia, COPD exacerbation, which turned had a couple admissions for less and history of smoking. She into a pneumonia with significant severe COPD exacerbations. She was recently diagnosed with shortness of breath. She was was eventually placed on home emphysema/COPD. She’s coming admitted to the hospital. She was oxygen, and then about 2 months Sample Case in for a routine follow-up for her sick enough to require BIPAP and ago her illness seemed to progress. Progression hypertension was in the ICU. Eventually, she You talk more, and it becomes clear with her daughter. recovered and was discharged that she doesn’t want to have to home. She is now in your office for go back to the hospital if it isn’t routine follow-up. necessary. She really prefers to stay at home. • Build trusting and respectful • Continue to build trusting, • Rely on the trusting, respectful relationships respectful relationships relationships that were built • Learn about the patient as a • Continue to learn more about the • Keep the focus on the patient person patient as a person as a person • Establish a surrogate decision • Ensure a good understanding • Ensure a good understanding of Conversation maker of diagnosis, prognosis, and diagnosis, prognosis, and treat- Goals • Promote patient-surrogate-family treatment options ment options before introducing conversations • Anticipate emergencies and make hospice a plan when appropriate • Continue to hope for the best, but • Promote patient-surrogate-family prepare for when things don’t go conversations well • Normalize the conversation • Talk about “what matters most” “You have been in and out of • Try starting it after family history “Can you tell me your the hospital quite a bit. How “Have you ever thought who would understanding of what happened in has that been?” speak for you if you couldn’t speak the hospital?” “How do you feel about your for yourself? Is it ok if we talk about “What was that like for you?” quality of life?” that?” “How are you doing now?” “Given everything that has • If they already have an advance happened, what are you hoping directive (AD): “If surrogate decision making was for?” needed, how was that?” Examples of “May I see it? What does it say?” “Unfortunately, we don’t have What to Say • If they do not have an AD: • Identify the values that guided any more treatments to help your decision making, i.e., “what lungs get better.” “Can I offer you some tools to start mattered most” thinking about it?” “It seems to me what matters most to you is to [stay out of the hospital, control your symptoms at home, and make the most of each day OR stay out of the hospital but continue to receive treatment] and I think [hospice OR home care] is the best way of doing that.”

New Codes from CMS • Use 99497 if you meet or exceed 16 minutes • Use 99497 + 99498 if you meet or exceed 46 minutes Documentation Requirements • If medical management billing is based on medical • Total time in minutes decision making, then you can bill as you normally Billing • Patient/surrogate/family “given opportunity to decline” would in that scenario. On top of that, you should also Details • Details of content bill based on time spent for ACP. • Attending MDs and DOs, as well as NPs and PAs • If instead you are billing for the medical management (i.e., those who are authorized to independently bill based on time, you should be sure you do not double Medicare for CPT services), are the only providers count the time spent on the advance care planning who can use these codes. conversation. Want more information? Visit ihi.org/CMSpayment The information contained in this document is based on our best understanding of the new reimbursement codes. It is your responsibility to check with your local billing expert before using the new codes. Please review our full disclaimer of warranties and liability at http://www.ihi.org/pages/termsofuse.aspx CAPC Website Update, October 2019 How to Create a CAPC User Account

Creating an Account 1. Visit CAPC.org and click “Create Account” in the upper right corner.

2. Select your “Organization” from the dropdown list of the organizations – you can start typing it in.

3. Enter your work email address, create a password, answer all questions, and agree to the terms.

4. Check your email Inbox for a verification email from [email protected] and click the link provided to finalize set-up. If the email does not show in your work inbox, please check junk/spam folder.

Accessing the CAPC Website 5. Click “Login” in the upper right corner on CAPC.org on all subsequent visits. Or, wherever you encounter members-only content (identified with a lock icon) click “Login”.

For assistance setting up your account or accessing courses, contact [email protected]

6. For a quick tour of CAPC.org and learn where all the different tools, resources, and courses lie, click on this four-minute video.

For assistance with content, please email [email protected] CLINICAL STATEMENTS AND GUIDELINES e137 August 8, 2017 August 8, © 2017 by the American College Foundation, the of Cardiology American Heart Association, Inc., Society of and the Heart Failure America Key Words: AHA Scientific Key Words: Statements ◼ angioedema ◼ angiotensin-converting enzyme inhibitors ◼ angiotensin receptor blockers ◼ angiotensin receptor- neprilysin inhibitor ◼ beta blockers ◼ ferric carboxymaltose ◼ focused update ◼ heart failure deficiency ◼ hypertension ◼ iron peptides ◼ ivabradine ◼ natriuretic peptide biomarker ◼ natriuretic ◼ sleep apnea Developed in Collaboration with Developed in Collaboration the American Academy of Family Physicians, American College of Chest Physicians, and International - Society for Heart and Lung Trans plantation Members, Force ACC/AHA Task see page e150 ‡ ¶ * ‡ * * † ** * ¶ * ‡ * Vice Chair , MD, FACC , MD* , MD, FACC# MD, MSPH, FACC , , MD, FACC, FAHA‡ MD, FACC, MD, FACC, FHFSA MD, FACC, § MD, MPH, MBA, FACC , MD, FACC, FAHA, FHFSA FAHA, MD, FACC, MD, MPH, FACC†† , MD, FACC, FAHA, FHFSA FAHA, MD, FACC, ║ MD, FAHA , , , MD, MSc, FACC, FAHA, FHFSA‡ FAHA, MD, MSc, FACC, , PhD, RN, ACNS-BC, FAHA, FHFSA¶ PhD, RN, ACNS-BC, FAHA, , MD, PhD, FACC, FAHA MD, PhD, FACC, , , , MD, FACC, FAHA MD, FACC, , MD, MBA, MPH, FACC, FAHA MD, MBA, MPH, FACC, ,

ISHLT Representative. ¶HFSA Representative. #CHEST Representative. **ACC/AHA Task Force on Force Task **ACC/AHA Representative. #CHEST Representative. ¶HFSA Representative. ISHLT

Cheryl Westlake Biykem Bozkurt Javed Butler Steven M. Hollenberg Michael M. Givertz Donald E. Casey Jr A. Masoudi Frederick Lynne Warner Stevenson Warner Lynne Gerasimos S. Filippatos JoAnn Lindenfeld Pamela N. Peterson Mark H. Drazner Monica M. Colvin Gregg C. Fonarow Gregg 2017;136:e137–e161. DOI: 10.1161/CIR.0000000000000509 DOI: Circulation. 2017;136:e137–e161. *Writing group members are required to recuse themselves from voting on sections to which their themselves from to recuse required members are group *Writing 1 for detailed information. †ACC/AHA see Appendix relationships with industry may apply; specific §ACP Representative. on Clinical Practice Guidelines Liaison. ‡ACC/AHA Representative. Force Task ‖ ††AAFP Representative. Representative. Performance Measures Jessup CW, that this document be cited as follows: Yancy The American Heart Association requests GC, Givertz Colvin MM, Drazner MH, Filippatos GS, Fonarow M, Bozkurt B, Butler J, Casey DE Jr, Westlake LW, McBride PE, Peterson PN, Stevenson MM, Hollenberg SM, Lindenfeld J, Masoudi FA, of C. 2017 ACC/AHA/HFSA focused update of the 2013 ACCF/AHA guideline for the management Force Heart Association Task of the American College of Cardiology/American a report heart failure: 2017;136:e137– Society of America. Circulation. on Clinical Practice Guidelines and the Heart Failure e161. DOI: 10.1161/CIR.0000000000000509. Patrick E. McBride WRITING GROUP MEMBERS* WRITING GROUP FHFSA, Chair FAHA, MD, MSc, MACC, Yancy, Clyde W. Mariell Jessup of Heart Failure of Heart Heart Cardiology/American the American College of A Report of and the Heart Failure on Clinical Practice Guidelines Force Task Association Society of America 2013 ACCF/AHA Guideline for the Management Guideline ACCF/AHA 2013 2017 ACC/AHA/HFSA Focused Update of the of the Update Focused ACC/AHA/HFSA 2017 ACC/AHA/HFSA FOCUSED UPDATE FOCUSED ACC/AHA/HFSA

Downloaded from http://ahajournals.org by on August 12, 2021 Yancy et al

TABLE OF CONTENTS Intended Use Preamble ��������������������������������������������������������������������� e138 Practice guidelines provide recommendations applica- 1. Introduction ������������������������������������������������������������� e139 ble to patients with or at risk of developing cardiovas- 1.1. Methodology and Evidence Review ����������������� e140 cular disease. The focus is on medical practice in the 1.2. Organization of the Writing Group ������������������� e141 United States, but guidelines developed in collabora- 1.3. Document Review and Approval ����������������������� e141 tion with other organizations may have a global impact. 6. Initial and Serial Evaluation of the HF Patient ����������� e141 6.3. Biomarkers ������������������������������������������������������� e141 Although guidelines may be used to inform regulatory 6.3.1. Biomarkers for Prevention: or payer decisions, their intent is to improve patients’ Recommendation ��������������������������������� e142 quality of care and align with patients’ interests. Guide- 6.3.2. Biomarkers for Diagnosis: lines are intended to define practices meeting the Recommendation ��������������������������������� e142 needs of patients in most, but not all, circumstances 6.3.3. Biomarkers for Prognosis or Added and should not replace clinical judgment. Risk Stratification: Recommendations �������� e142 7. Treatment of Stages A to D ������������������������������������� e144 7.3. Stage C ����������������������������������������������������������� e144 Clinical Implementation 7.3.2. Pharmacological Treatment for Stage Guideline recommended management is effective only C HF With Reduced Ejection Fraction: Recommendations ��������������������������������� e144 when followed by healthcare providers and patients. 7.3.2.10. Renin-Angiotensin System Adherence to recommendations can be enhanced by Inhibition With Angiotensin- shared decision making between healthcare providers Converting Enzyme Inhibitor and patients, with patient engagement in selecting or Angiotensin Receptor Blocker interventions based on individual values, preferences, or ARNI: Recommendations ��� e144 and associated conditions and comorbidities. 7.3.2.11. Ivabradine: Recommendation ����e145 7.3.3. Pharmacological Treatment for Stage C HFpEF: Recommendations ��������������������� e146 Methodology and Modernization 9. Important Comorbidities in HF ��������������������������������� e147 The ACC/AHA Task Force on Clinical Practice Guidelines 9.2. Anemia: Recommendations ����������������������������� e147 9.5. Hypertension (New Section) ����������������������������� e149 (Task Force) continuously reviews, updates, and modi- Downloaded from http://ahajournals.org by on August 12, 2021 9.5.1. Treating Hypertension to Reduce the fies guideline methodology on the basis of published Incidence of HF: Recommendation ��������� e149 standards from organizations including the Institute of 9.5.2. Treating Hypertension in Stage C Medicine1,2 and on the basis of internal reevaluation. HFrEF: Recommendation ����������������������� e149 Similarly, the presentation and delivery of guidelines are 9.5.3. Treating Hypertension in Stage C reevaluated and modified on the basis of evolving tech- HFpEF: Recommendation ����������������������� e149 nologies and other factors to facilitate optimal dissemi- 9.6. Sleep-Disordered Breathing: nation of information at the point of care to healthcare Recommendations ������������������������������������������� e149 professionals. Given time constraints of busy healthcare References ������������������������������������������������������������������� e151 Appendix 1. Author Relationships With Industry providers and the need to limit text, the current guideline and Other Entities (Relevant) ��������������������� e156 format delineates that each recommendation be support- Appendix 2. Reviewer Relationships With Industry ed by limited text (ideally, <250 words) and hyperlinks to and Other Entities (Comprehensive) ��������� e158 supportive evidence summary tables. Ongoing efforts to Appendix 3. Abbreviations ������������������������������������������� e161 further limit text are underway. Recognizing the impor- tance of cost-value considerations in certain guidelines, when appropriate and feasible, an analysis of the value PREAMBLE of a drug, device, or intervention may be performed in Since 1980, the American College of Cardiology (ACC) accordance with the ACC/AHA methodology.3 and American Heart Association (AHA) have translat- To ensure that guideline recommendations remain ed scientific evidence into clinical practice guidelines current, new data are reviewed on an ongoing basis, (guidelines) with recommendations to improve cardio- with full guideline revisions commissioned in approxi- vascular health. These guidelines, which are based on mately 6-year cycles. Publication of new, potentially systematic methods to evaluate and classify evidence, practice-changing study results that are relevant to an provide a cornerstone for quality cardiovascular care. existing or new drug, device, or management strategy The ACC and AHA sponsor the development and pub- will prompt evaluation by the Task Force, in consultation lication of guidelines without commercial support, and with the relevant guideline writing committee, to de- members of each organization volunteer their time to termine whether a focused update should be commis- the writing and review efforts. Guidelines are official sioned. For additional information and policies regard- policy of the ACC and AHA. ing guideline development, we encourage readers to

e138 August 8, 2017 Circulation. 2017;136:e137–e161. DOI: 10.1161/CIR.0000000000000509 2017 ACC/AHA/HFSA Heart Failure Focused Update CLINICAL STATEMENTS AND GUIDELINES consult the ACC/AHA guideline methodology manual4 Guideline-Directed Management and other methodology articles.5–8 and Therapy The term guideline-directed management and therapy Selection of Writing Committee Members (GDMT) encompasses clinical evaluation, diagnostic The Task Force strives to avoid bias by selecting experts testing, and pharmacological and procedural treat- ments. For these and all recommended drug treat-

from a broad array of backgrounds. Writing commit- tee members represent different geographic regions, ment regimens, the reader should confirm the dosage sexes, ethnicities, races, intellectual perspectives/biases, by reviewing product insert material and evaluate the and scopes of clinical practice. The Task Force may also treatment regimen for contraindications and interac- invite organizations and professional societies with re- tions. The recommendations are limited to drugs, de- lated interests and expertise to participate as partners, vices, and treatments approved for clinical use in the collaborators, or endorsers. United States.

Relationships With Industry and Other Class of Recommendation and Level Entities of Evidence The ACC and AHA have rigorous policies and meth- The Class of Recommendation (COR) indicates the ods to ensure that guidelines are developed without strength of the recommendation, encompassing the bias or improper influence. The complete relationships estimated magnitude and certainty of benefit in pro- with industry and other entities (RWI) policy can be portion to risk. The Level of Evidence (LOE) rates the found online. Appendix 1 of the current document quality of scientific evidence that supports the inter- lists writing committee members’ relevant RWI. For the vention on the basis of the type, quantity, and con- purposes of full transparency, writing committee mem- sistency of data from clinical trials and other sources 4–6 bers’ comprehensive disclosure information is available (Table 1). online. Comprehensive disclosure information for the Glenn N. Levine, MD, FACC, FAHA Task Force is also available online. Chair, ACC/AHA Task Force on Clinical Practice Guidelines Downloaded from http://ahajournals.org by on August 12, 2021 Evidence Review and Evidence Review Committees 1. INTRODUCTION When developing recommendations, the writing com- The purpose of this focused update is to update the mittee uses evidence-based methodologies that are “2013 ACCF/AHA Guideline for the Management of based on all available data.4–7 Literature searches focus Heart Failure”9 (2013 HF guideline) in areas in which on randomized controlled trials (RCTs) but also include new evidence has emerged since its publication. For registries, nonrandomized comparative and descriptive this update and future heart failure (HF) guidelines, the studies, case series, cohort studies, systematic reviews, Heart Failure Society of America (HFSA) has partnered and expert opinion. Only key references are cited. with the ACC and AHA to provide coordinated guid- An independent evidence review committee (ERC) ance on the management of HF. is commissioned when there are 1 or more questions The scope of the focused update includes revision deemed of utmost clinical importance that merit formal to the sections on biomarkers; new therapies indicated systematic review. This systematic review will strive to de- for stage C HF with reduced ejection fraction (HFrEF); termine which patients are most likely to benefit from a updates on HF with preserved ejection fraction (HFpEF); drug, device, or treatment strategy and to what degree. new data on important comorbidities, including sleep Criteria for commissioning an ERC and formal systematic apnea, anemia, and hypertension; and new insights review include: a) the absence of a current authoritative into the prevention of HF. systematic review, b) the feasibility of defining the ben- This focused update represents the second part of a efit and risk in a time frame consistent with the writing 2-stage publication; with the first part having been pub- of a guideline, c) the relevance to a substantial number lished as the “2016 ACC/AHA/HFSA Focused Update of patients, and d) the likelihood that the findings can on New Pharmacological Therapy for Heart Failure,”10 be translated into actionable recommendations. ERC which introduced guidance on new therapies, specifi- members may include methodologists, epidemiologists, cally for the use of an angiotensin receptor–neprilysin healthcare providers, and biostatisticians. When a formal inhibitor (ARNI) (valsartan/sacubitril) and a sinoatrial systematic review has been commissioned, the recom- node modulator (ivabradine). That focused update was mendations developed by the writing committee on the published concurrently with the European Society of basis of the systematic review are marked with“SR”. Cardiology’s complete guideline, “2016 ESC Guidelines

Circulation. 2017;136:e137–e161. DOI: 10.1161/CIR.0000000000000509 August 8, 2017 e139 Yancy et al

Table 1. Applying Class of Recommendation and Level of Evidence to Clinical Strategies, Interventions, Treatments, or Diagnostic Testing in Patient Care* (Updated August 2015) Downloaded from http://ahajournals.org by on August 12, 2021

for the Diagnosis and Treatment of Acute and Chronic ACC, AHA, and European Society of Cardiology and Heart Failure.”11 other scientific meetings and that were published in peer-reviewed format from April 2013 through Novem- ber 2016. The evidence is summarized in tables in the 1.1. Methodology and Evidence Review Online Data Supplement. All recommendations (new, To identify key data that influence guideline recommen- modified, and unchanged) for each clinical section are dations, the Task Force and members of the 2013 HF included to provide a comprehensive assessment. The guideline writing committee reviewed clinical trials that text explains new and modified recommendations, were presented at the annual scientific meetings of the whereas recommendations from the previous guideline

e140 August 8, 2017 Circulation. 2017;136:e137–e161. DOI: 10.1161/CIR.0000000000000509 2017 ACC/AHA/HFSA Heart Failure Focused Update CLINICAL STATEMENTS AND GUIDELINES that have been deleted or superseded no longer ap- severity of HF. In general, both natriuretic peptide bio- pear. Please consult the full-text version of the 2013 marker values track similarly, and either can be used HF guideline9 for text and evidence tables supporting in patient care settings as long as their respective ab- the unchanged recommendations and for clinical areas solute values and cutpoints are not used interchange- not addressed in this focused update. Individual recom- ably. Notably, BNP, but not NT-proBNP, is a substrate mendations in this focused update will be incorporated for neprilysin. Therefore, ARNI increases BNP levels12 13 into the full-text guideline in the future. Recommenda- but not NT-proBNP levels. Note that the type of natri- tions from the prior guideline that remain current have uretic peptide assay that has been performed must be been included for completeness, but the LOE reflects considered during interpretation of natriuretic peptide the COR/LOE system used when the recommendations biomarker levels in patients on ARNI. In 2 studies with were initially developed. New and modified recommen- ARNI, NT-proBNP levels were reduced,12,14 with the dations in this focused update reflect the latest COR/ reduction in 1 study being associated with improved LOE system, in which LOE B and C are subcategorized clinical outcomes.12 for greater specificity.4–6 The section numbers corre- A substantial evidence base exists that supports the spond to the full-text guideline sections. use of natriuretic peptide biomarkers to assist in the di- agnosis or exclusion of HF as a cause of symptoms (eg, dyspnea, weight gain) in the setting of chronic ambula- 1.2. Organization of the Writing Group tory HF15–21 or in the setting of acute care with decom- For this focused update, representative members of pensated HF,22–30 especially when the cause of dyspnea the 2013 HF guideline writing committee were invit- is unclear. The role of natriuretic peptide biomarkers in ed to participate. They were joined by additional in- population screening to detect incident HF is emerg- vited members to form a new writing group, which ing.31–37 Elevated plasma levels of natriuretic peptide is referred to as the 2017 HF focused update writing biomarkers are associated with a wide variety of cardiac group. Members were required to disclose all RWI rel- and noncardiac causes (Table 2).38–42 Obesity may be as- evant to the data under consideration. The group was sociated with lower natriuretic peptide concentrations, composed of experts representing general cardiolo- and this may modestly reduce diagnostic sensitivity in gists, HF and transplantation specialists, electrophysi- morbidly obese patients.42 ologists, pharmacists, and general internists. The 2017 Downloaded from http://ahajournals.org by on August 12, 2021 HF focused update writing group included representa- Table 2. Selected Potential Causes of Elevated tives from the ACC, AHA, and HFSA, as well as the Natriuretic Peptide Levels38–41 American Academy of Family Physicians, American Cardiac College of Chest Physicians, American College of Phy- sicians, and International Society for Heart and Lung HF, including RV syndromes Transplantation. Acute coronary syndromes Heart muscle disease, including LVH 1.3. Document Review and Approval Valvular heart disease Pericardial disease The focused update was reviewed by 2 official reviewers Atrial fibrillation each nominated by the ACC, AHA, and HFSA; 1 reviewer each from the American Academy of Family Physicians, Myocarditis American College of Chest Physicians, and International Cardiac surgery Society for Heart and Lung Transplantation; and 19 in- Cardioversion dividual content reviewers. Reviewers’ RWI information Toxic-metabolic myocardial insults, including cancer chemotherapy is published in this document (Appendix 2). Noncardiac This document was approved for publication by the Advancing age governing bodies of the ACC, AHA, and HFSA. Anemia Renal failure 6. INITIAL AND SERIAL EVALUATION Pulmonary: obstructive sleep apnea, severe pneumonia OF THE HF PATIENT Pulmonary hypertension Critical illness 6.3. Biomarkers Bacterial sepsis Assays for BNP (B-type natriuretic peptide) and NT- Severe burns proBNP (N-terminal pro-B-type natriuretic peptide), Modified from Table 8 of the 2013 HF guideline.9 which are both natriuretic peptide biomarkers, have HF, indicates heart failure; LVH, left ventricular hypertrophy; and RV, right been used increasingly to establish the presence and ventricular.

Circulation. 2017;136:e137–e161. DOI: 10.1161/CIR.0000000000000509 August 8, 2017 e141 Yancy et al

Because of the absence of clear and consistent evi- mation, oxidative stress, vascular dysfunction, and myo- dence for improvement in mortality and cardiovascular cardial and matrix remodeling, have been implicated in outcomes,43–62 there are insufficient data to inform spe- HF. 68–71 Biomarkers of myocardial fibrosis, soluble ST2 cific guideline recommendations related to natriuretic receptor, and galectin-3 are predictive of hospitaliza- peptide–guided therapy or serial measurements of BNP tion and death and may provide incremental prognos- or NT-proBNP levels for the purpose of reducing hospi- tic value over natriuretic peptide levels in patients with talization or deaths in the present document. HF. 72–74 Strategies that combine multiple biomarkers Like natriuretic peptides, cardiac troponin levels may may ultimately prove beneficial in guiding HF therapy be elevated in the setting of chronic or acute decom- in the future, but multicenter studies with larger deri- pensated HF, suggesting myocyte injury or necrosis.63 vation and validation cohorts are needed.75,76 Several Troponins I and T respond similarly for acute coronary emerging biomarkers await validation with well-defined syndromes and acute decompensated HF. Elevations in outcome measures and prognostic accuracy before they either troponin I or T levels in the setting of acute HF can reach the clinical arena.77–84 are of prognostic significance and must be interpreted This section categorizes the role of biomarkers into in the clinical context.64 prevention, diagnosis, prognosis, and added risk strati- In addition to natriuretic peptides and troponins,65–67 fication to clarify evidence-based objectives of their use multiple other biomarkers, including those of inflam- in clinical practice.

6.3.1 Biomarkers for Prevention: Recommendation Biomarkers: Recommendation for Prevention of HF COR LOE Recommendation Comment/Rationale IIa B-R For patients at risk of developing HF, natriuretic peptide biomarker– NEW: New data suggest that natriuretic based screening followed by team-based care, including a peptide biomarker screening and early See Online Data cardiovascular specialist optimizing GDMT, can be useful to prevent the intervention may prevent HF. Supplements A and B. development of left ventricular dysfunction (systolic or diastolic) or new-onset HF.85,86 In a large-scale unblinded single-center study (STOP-HF [The St Vincent’s Screening to Prevent Heart Failure]),85 patients at risk of HF (identified by the presence of hypertension, diabetes mellitus, or known vascular disease [eg, stage A HF]), but without established left ventricular systolic dysfunction Downloaded from http://ahajournals.org by on August 12, 2021 or symptomatic HF at baseline, were randomly assigned to receive screening with BNP testing or usual primary care. Intervention-group participants with BNP levels of ≥50 pg/mL underwent echocardiography and were referred to a cardiovascular specialist who decided on further investigation and management. All patients received further coaching by a specialist nurse who emphasized individual risk and the importance of adherence to medication and healthy lifestyle behaviors. BNP-based screening reduced the composite endpoint of asymptomatic left ventricular dysfunction (systolic or diastolic) with or without newly diagnosed HF.85 Similarly, in another small, single-center RCT, accelerated up-titration of renin-angiotensin-aldosterone system antagonists and beta blockers reduced cardiac events in patients with diabetes mellitus and elevated NT-proBNP levels but without cardiac disease at baseline.86 Developing a standardized strategy to screen and intervene in patients at risk of HF can be difficult because of different definitions of HF risk, heterogeneity of prevalence in different populations, variable duration until clinical HF or left ventricular dysfunction develops, and variable interventions for risk factor modification or treatment. Further studies are needed to determine cost-effectiveness and risk of such screening, as well as its impact on quality of life (QoL) and mortality rate.

6.3.2 Biomarkers for Diagnosis: Recommendation Biomarkers: Recommendation for Diagnosis COR LOE Recommendation Comment/Rationale I A In patients presenting with dyspnea, measurement of natriuretic MODIFIED: 2013 acute and chronic peptide biomarkers is useful to support a diagnosis or exclusion recommendations have been combined into a See Online Data of HF.15–24,28–30 diagnosis section. Supplements A and B. Natriuretic peptide biomarker testing in the setting of chronic ambulatory HF provides incremental diagnostic value to clinical judgment, especially when the etiology of dyspnea is unclear.15–21 In emergency settings, natriuretic peptide biomarker levels usually have higher sensitivity than specificity and may be more useful for ruling out than ruling in HF.20 Although lower values of natriuretic peptide biomarkers exclude the presence of HF, and higher values have reasonably high positive predictive value to diagnose HF, clinicians should be aware that elevated plasma levels for both natriuretic peptides have been associated with a wide variety of cardiac and noncardiac causes (Table 2).38–41

6.3.3 Biomarkers for Prognosis or Added Risk Stratification: Recommendations Biomarkers: Recommendations for Prognosis COR LOE Recommendations Comment/Rationale Measurement of BNP or NT-proBNP is useful for establishing prognosis 2013 recommendation remains current. I A or disease severity in chronic HF.16,87–92

e142 August 8, 2017 Circulation. 2017;136:e137–e161. DOI: 10.1161/CIR.0000000000000509 2017 ACC/AHA/HFSA Heart Failure Focused Update CLINICAL STATEMENTS AND GUIDELINES Biomarkers: Recommendations for Prognosis (Continued) COR LOE Recommendations Comment/Rationale I A Measurement of baseline levels of natriuretic peptide biomarkers and/ MODIFIED: Current recommendation or cardiac troponin on admission to the hospital is useful to establish a emphasizes that it is admission levels of See Online Data prognosis in acutely decompensated HF.27,93–100 natriuretic peptide biomarkers that are useful. Supplements A and B. Higher levels of natriuretic peptide biomarkers on admission are usually associated with greater risk for clinical outcomes, including all-cause and cardiovascular mortality, morbidity, and composite outcomes, across different time intervals in patients with decompensated HF.20,27,29,93–101 Similarly, abnormal levels of circulating cardiac troponin are commonly found in patients with acute decompensated HF, often without obvious myocardial ischemia or underlying coronary artery disease (CAD), and this is associated with worse clinical outcomes and higher risk of death.95,99,102,103 Studies have demonstrated incremental prognostic value of these biomarkers to standard approaches of cardiovascular disease risk assessment.29,95 However, there were differences in the risk prediction models, assay cutpoints, and lengths of follow-up.29 Furthermore, not all patients may need biomarker measurement for prognostication, especially if they already have advanced HF with established poor prognosis or persistently elevated levels of biomarkers in former settings. Therefore, assays of natriuretic peptide biomarkers for incremental prognostication should not preclude good clinical judgment; an individualized approach to each patient is paramount. IIa B-NR During a HF hospitalization, a predischarge natriuretic peptide level can NEW: Current recommendation reflects new be useful to establish a postdischarge prognosis.93,96,104–113 observational studies. See Online Data Supplements A and B. Predischarge natriuretic peptide biomarker levels and the relative change in levels during hospital treatment are strong predictors of the risk of death or hospital readmission for HF.93,96,104–113 Several studies have suggested that predischarge natriuretic peptide biomarker levels had higher reclassification and discrimination value than clinical variables in predicting outcomes.96,106,108–111 Patients with higher predischarge levels and patients who do not have a decrease in natriuretic peptide biomarker levels during hospitalization have worse outcomes.96,106,108–111 Although observational or retrospective studies have suggested that patients with natriuretic peptide biomarker reduction had better outcomes than those without any changes or with a biomarker rise,93,107,112,113 targeting a certain threshold, value, or relative change in these biomarker levels during hospitalization may not be practical or safe for every patient and has not been tested in a prospective large-scale trial. Clinical assessment and adherence to GDMT should be the emphasis, and the prognostic value of a predischarge value or relative changes does not imply the necessity for serial and repeated biomarker measurements during hospitalization. IIb B-NR In patients with chronic HF, measurement of other clinically available MODIFIED: 2013 recommendations have been tests, such as biomarkers of myocardial injury or fibrosis, may be combined into prognosis section, resulting in See Online Data considered for additive risk stratification.27,95,98,99,103,114–119 LOE change from A to B-NR. Supplements A and B. Biomarkers of myocardial fibrosis (eg, soluble ST2 receptor, galectin-3, high-sensitivity cardiac troponin, and others) are predictive of hospitalization and death in patients with HF and also are additive to natriuretic peptide biomarker levels in their prognostic value.117,119–126 A combination of biomarkers may ultimately prove to be more informative than single biomarkers.127 Downloaded from http://ahajournals.org by on August 12, 2021

Figure 1. Biomarkers Indications for Use. Colors correspond to COR in Table 1. *Other biomarkers of injury or fibrosis include soluble ST2 receptor, galectin-3, and high- sensitivity troponin. ACC indicates American College of Cardiology; AHA, American Heart Association; ADHF, acute decom- pensated heart failure; BNP, B-type natriuretic peptide; COR, Class of Recommendation; ED, emergency department; HF, heart failure; NT-proBNP, N-terminal pro-B-type natriuretic peptide; NYHA, New York Heart Association; and pts, patients.

Circulation. 2017;136:e137–e161. DOI: 10.1161/CIR.0000000000000509 August 8, 2017 e143 Yancy et al

7. Treatment of Stages A to D 7.3. Stage C 7.3.2. Pharmacological Treatment for Stage C HF With Reduced Ejection Fraction: Recommendations (See Figure 2 and Table 3). 7.3.2.10. Renin-Angiotensin System Inhibition With Angiotensin-Converting Enzyme Inhibitor or Angiotensin Receptor Blocker or ARNI: Recommendations Recommendations for Renin-Angiotensin System Inhibition With ACE Inhibitor or ARB or ARNI COR LOE Recommendations Comment/Rationale ACE-I: A The clinical strategy of inhibition of the renin-angiotensin system with NEW: New clinical trial data prompted ACE inhibitors (Level of Evidence: A),128–133 OR ARBs (Level of Evidence: clarification and important updates. ARB: A A),134–137 OR ARNI (Level of Evidence: B-R)138 in conjunction with evidence- I based beta blockers,9,139,140 and aldosterone antagonists in selected ARNI: B-R patients,141,142 is recommended for patients with chronic HFrEF to reduce morbidity and mortality. Angiotensin-converting enzyme (ACE) inhibitors reduce morbidity and mortality in heart failure with reduced ejection fraction (HFrEF). Randomized controlled trials (RCTs) clearly establish the benefits of ACE inhibition in patients with mild, moderate, or severe symptoms of HF and in patients with or without coronary artery disease.128–133 ACE inhibitors can produce angioedema and should be given with caution to patients with low systemic blood pressures, renal insufficiency, or elevated serum potassium. ACE inhibitors also inhibit kininase and increase levels of bradykinin, which can induce cough but also may contribute to their beneficial effect through vasodilation. Angiotensin receptor blockers (ARBs) were developed with the rationale that angiotensin II production continues in the presence of ACE inhibition, driven through alternative enzyme pathways. ARBs do not inhibit kininase and are associated with a much See Online Data lower incidence of cough and angioedema than ACE inhibitors; but like ACE inhibitors, ARBs should be given with caution to Supplements 1, 2, patients with low systemic blood pressure, renal insufficiency, or elevated serum potassium. Long-term therapy with ARBs produces 18-20. hemodynamic, neurohormonal, and clinical effects consistent with those expected after interference with the renin-angiotensin system and have been shown in RCTs134–137 to reduce morbidity and mortality, especially in ACE inhibitor–intolerant patients. In ARNI, an ARB is combined with an inhibitor of neprilysin, an enzyme that degrades natriuretic peptides, bradykinin, adrenomedullin, and other vasoactive peptides. In an RCT that compared the first approved ARNI, valsartan/sacubitril, with enalapril in symptomatic patients with HFrEF tolerating an adequate dose of either ACE inhibitor or ARB, the ARNI reduced the composite endpoint of cardiovascular death or HF hospitalization significantly, by 20%.138 The benefit was seen to a similar extent for both death and HF hospitalization and was consistent across subgroups. The use of ARNI is associated with the risk of hypotension and Downloaded from http://ahajournals.org by on August 12, 2021 renal insufficiency and may lead to angioedema, as well. The use of ACE inhibitors is beneficial for patients with prior or current 2013 recommendation repeated for clarity in I ACE-I: A symptoms of chronic HFrEF to reduce morbidity and mortality.128–133,143 this section. ACE inhibitors have been shown in large RCTs to reduce morbidity and mortality in patients with HFrEF with mild, moderate, or severe symptoms of HF, with or without coronary artery disease.128–133 Data suggest that there are no differences among available ACE inhibitors in their effects on symptoms or survival.143 ACE inhibitors should be started at low doses and titrated upward to doses shown to reduce the risk of cardiovascular events in clinical trials. ACE inhibitors can produce angioedema and should be given with caution to patients with low systemic blood pressures, renal insufficiency, or elevated serum potassium (>5.0 mEq/L). See Online Data Angioedema occurs in <1% of patients who take an ACE inhibitor, but it occurs more frequently in blacks and women.144 Patients Supplement 18. should not be given ACE inhibitors if they are pregnant or plan to become pregnant. ACE inhibitors also inhibit kininase and increase levels of bradykinin, which can induce cough in up to 20% of patients but also may contribute to beneficial vasodilation. If maximal doses are not tolerated, intermediate doses should be tried; abrupt withdrawal of ACE inhibition can lead to clinical deterioration and should be avoided. Although the use of an ARNI in lieu of an ACE inhibitor for HFrEF has been found to be superior, for those patients for whom ARNI is not appropriate, continued use of an ACE inhibitor for all classes of HFrEF remains strongly advised. The use of ARBs to reduce morbidity and mortality is recommended 2013 recommendation repeated for clarity in I ARB: A in patients with prior or current symptoms of chronic HFrEF who are this section. intolerant to ACE inhibitors because of cough or angioedema.134–137,145,146 ARBs have been shown to reduce mortality and HF hospitalizations in patients with HFrEF in large RCTs.134–137 Long-term therapy with ARBs in patients with HFrEF produces hemodynamic, neurohormonal, and clinical effects consistent with those expected after interference with the renin-angiotensin system.145,146 Unlike ACE inhibitors, ARBs do not inhibit kininase and are associated with a much lower incidence of cough and angioedema, although kininase inhibition by ACE inhibitors may produce beneficial vasodilatory effects. Patients intolerant to ACE inhibitors because of cough or angioedema should be started on ARBs; patients already tolerating See Online Data ARBs for other indications may be continued on ARBs if they subsequently develop HF. ARBs should be started at low doses Supplements 2 and 19. and titrated upward, with an attempt to use doses shown to reduce the risk of cardiovascular events in clinical trials. ARBs should be given with caution to patients with low systemic blood pressure, renal insufficiency, or elevated serum potassium (>5.0 mEq/L). Although ARBs are alternatives for patients with ACE inhibitor–induced angioedema, caution is advised because some patients have also developed angioedema with ARBs. Head-to-head comparisons of an ARB versus ARNI for HF do not exist. For those patients for whom an ACE inhibitor or ARNI is inappropriate, use of an ARB remains advised.

e144 August 8, 2017 Circulation. 2017;136:e137–e161. DOI: 10.1161/CIR.0000000000000509 2017 ACC/AHA/HFSA Heart Failure Focused Update CLINICAL STATEMENTS AND GUIDELINES Recommendations for Renin-Angiotensin System Inhibition With ACE Inhibitor or ARB or ARNI (Continued) COR LOE Recommendations Comment/Rationale In patients with chronic symptomatic HFrEF NYHA class II or III NEW: New clinical trial data necessitated this I ARNI: B-R who tolerate an ACE inhibitor or ARB, replacement by an ARNI is recommendation. recommended to further reduce morbidity and mortality.138 Benefits of ACE inhibitors with regard to decreasing HF progression, hospitalizations, and mortality rate have been shown

consistently for patients across the clinical spectrum, from asymptomatic to severely symptomatic HF. Similar benefits have been shown for ARBs in populations with mild-to-moderate HF who are unable to tolerate ACE inhibitors. In patients with mild-to- moderate HF (characterized by either 1) mildly elevated natriuretic peptide levels, BNP [B-type natriuretic peptide] >150 pg/mL or NT-proBNP [N-terminal pro-B-type natriuretic peptide] ≥600 pg/mL; or 2) BNP ≥100 pg/mL or NT-proBNP ≥400 pg/mL with a prior hospitalization in the preceding 12 months) who were able to tolerate both a target dose of enalapril (10 mg twice daily) and then subsequently an ARNI (valsartan/sacubitril; 200 mg twice daily, with the ARB component equivalent to valsartan 160 See Online Data mg), hospitalizations and mortality were significantly decreased with the valsartan/sacubitril compound compared with enalapril. Supplements 1 and 18. The target dose of the ACE inhibitor was consistent with that known to improve outcomes in previous landmark clinical trials.129 This ARNI has been approved for patients with symptomatic HFrEF and is intended to be substituted for ACE inhibitors or ARBs. HF effects and potential off-target effects may be complex with inhibition of the neprilysin enzyme, which has multiple biological targets. Use of an ARNI is associated with hypotension and a low-frequency incidence of angioedema. To facilitate initiation and titration, the approved ARNI is available in 3 doses that include a dose that was not tested in the HF trial; the target dose used in the trial was 97/103 mg twice daily.147 Clinical experience will provide further information about the optimal titration and tolerability of ARNI, particularly with regard to blood pressure, adjustment of concomitant HF medications, and the rare complication of angioedema.14 ARNI should not be administered concomitantly with ACE inhibitors or NEW: Available evidence demonstrates a III: Harm B-R within 36 hours of the last dose of an ACE inhibitor.148,149 potential signal of harm for a concomitant use of ACE inhibitors and ARNI. Oral neprilysin inhibitors, used in combination with ACE inhibitors can lead to angioedema and concomitant use is contraindicated and should be avoided. A medication that represented both a neprilysin inhibitor and an ACE inhibitor, omapatrilat, was studied See Online Data in both hypertension and HF, but its development was terminated because of an unacceptable incidence of angioedema148,149 and Supplement 3. associated significant morbidity. This adverse effect was thought to occur because both ACE and neprilysin break down bradykinin, which directly or indirectly can cause angioedema.149,150 An ARNI should not be administered within 36 hours of switching from or to an ACE inhibitor. ARNI should not be administered to patients with a history of NEW: New clinical trial data. III: Harm C-EO angioedema. Omapatrilat, a neprilysin inhibitor (as well as an ACE inhibitor and aminopeptidase P inhibitor), was associated with a higher Downloaded from http://ahajournals.org by on August 12, 2021 frequency of angioedema than that seen with enalapril in an RCT of patients with HFrEF. 148 In a very large RCT of hypertensive patients, omapatrilat was associated with a 3-fold increased risk of angioedema as compared with enalapril.149 Blacks and smokers were particularly at risk. The high incidence of angioedema ultimately led to cessation of the clinical development of N/A omapatrilat.151,152 In light of these observations, angioedema was an exclusion criterion in the first large trial assessing ARNI therapy in patients with hypertension153 and then in the large trial that demonstrated clinical benefit of ARNI therapy in HFrEF. 138 ARNI therapy should not be administered in patients with a history of angioedema because of the concern that it will increase the risk of a recurrence of angioedema. 7.3.2.11. Ivabradine: Recommendation Recommendation for Ivabradine COR LOE Recommendation Comment/Rationale Ivabradine can be beneficial to reduce HF hospitalization for patients NEW: New clinical trial data. with symptomatic (NYHA class II-III) stable chronic HFrEF (LVEF ≤35%) IIa B-R who are receiving GDEM*, including a beta blocker at maximum tolerated dose, and who are in sinus rhythm with a heart rate of 70 bpm or greater at rest.154–157

Ivabradine is a new therapeutic agent that selectively inhibits the If current in the sinoatrial node, providing heart rate reduction. One RCT demonstrated the efficacy of ivabradine in reducing the composite endpoint of cardiovascular death or HF hospitalization.155 The benefit of ivabradine was driven by a reduction in HF hospitalization. The study included patients with HFrEF (NYHA class II-IV, albeit with only a modest representation of NYHA class IV HF) and left ventricular ejection fraction (LVEF) ≤35%, in sinus rhythm with a resting heart rate of ≥70 beats per minute. Patients enrolled included a small number with paroxysmal See Online Data atrial fibrillation (<40% of the time) but otherwise in sinus rhythm and a small number experiencing ventricular pacing but with Supplement 4. a predominant sinus rhythm. Those with a myocardial infarction within the preceding 2 months were excluded. Patients enrolled had been hospitalized for HF in the preceding 12 months and were on stable GDEM* for 4 weeks before initiation of ivabradine therapy. The target of ivabradine is heart rate slowing (the presumed benefit of action), but only 25% of patients studied were on optimal doses of beta-blocker therapy.9,139,140,155 Given the well-proven mortality benefits of beta-blocker therapy, it is important to initiate and up titrate these agents to target doses, as tolerated, before assessing the resting heart rate for consideration of ivabradine initiation.155

*In other parts of the document, the term “GDMT” has been used to denote guideline-directed management and therapy. In this recommendation, however, the term “GDEM” has been used to denote this same concept in order to reflect the original wording of the recommendation that initially appeared in the “2016 ACC/AHA/HFSA Focused Update on New Pharmacological Therapy for Heart Failure: An Update of the 2013 ACCF/AHA Guideline for the Management of Heart Failure.”10

Circulation. 2017;136:e137–e161. DOI: 10.1161/CIR.0000000000000509 August 8, 2017 e145 Yancy et al Downloaded from http://ahajournals.org by on August 12, 2021 Figure 2. Treatment of HFrEF Stage C and D. Colors correspond to COR in Table 1. For all medical therapies, dosing should be optimized and serial assessment exercised. *See text for important treatment directions. †Hydral-Nitrates green box: The combination of ISDN/HYD with ARNI has not been robustly tested. BP response should be carefully monitored. ‡See 2013 HF guideline.9 §Participation in investigational studies is also appropriate for stage C, NYHA class II and III HF. ACEI indicates angiotensin-converting enzyme inhibitor; ARB, angiotensin receptor-blocker; ARNI, angiotensin receptor-neprilysin inhibitor; BP, blood pressure; bpm, beats per minute; C/I, contraindication; COR, Class of Recommendation; CrCl, creatinine clearance; CRT-D, cardiac resynchronization therapy–device; Dx, diagnosis; GDMT, guideline-directed management and therapy; HF, heart failure; HFrEF, heart failure with reduced ejection fraction; ICD, implantable cardioverter-defibrillator; ISDN/HYD, isosorbide dinitrate hydral-nitrates; K+, potassium; LBBB, left bundle-branch block; LVAD, left ventricular assist device; LVEF, left ventricular ejection fraction; MI, myocardial infarction; NSR, normal sinus rhythm; and NYHA, New York Heart Association.

7.3.3. Pharmacological Treatment for Stage C HFpEF: Recommendations Recommendations for Stage C HFpEF COR LOE Recommendations Comment/Rationale Systolic and diastolic blood pressure should be controlled in patients 2013 recommendation remains current. I B with HFpEF in accordance with published clinical practice guidelines to prevent morbidity.164,165 Diuretics should be used for relief of symptoms due to volume overload 2013 recommendation remains current. I C in patients with HFpEF. Coronary revascularization is reasonable in patients with CAD in whom 2013 recommendation remains current. IIa C symptoms (angina) or demonstrable myocardial ischemia is judged to be having an adverse effect on symptomatic HFpEF despite GDMT. Management of AF according to published clinical practice guidelines in 2013 recommendation remains current (Section IIa C patients with HFpEF is reasonable to improve symptomatic HF. 9.1 in the 2013 HF guideline). The use of beta-blocking agents, ACE inhibitors, and ARBs in patients 2013 recommendation remains current. IIa C with hypertension is reasonable to control blood pressure in patients with HFpEF.

e146 August 8, 2017 Circulation. 2017;136:e137–e161. DOI: 10.1161/CIR.0000000000000509 2017 ACC/AHA/HFSA Heart Failure Focused Update CLINICAL STATEMENTS AND GUIDELINES Recommendations for Stage C HFpEF (Continued) COR LOE Recommendations Comment/Rationale IIb B-R In appropriately selected patients with HFpEF (with EF ≥45%, elevated NEW: Current recommendation reflects new BNP levels or HF admission within 1 year, estimated glomerular filtration RCT data. See Online Data rate >30 mL/min, creatinine <2.5 mg/dL, potassium <5.0 mEq/L), Supplement C. aldosterone receptor antagonists might be considered to decrease hospitalizations.83,166,167

Mechanistic studies have suggested that mineralocorticoid receptor antagonists can improve measures of diastolic function in patients with HFpEF, possibly by a similar effect on remodeling.83,168 The TOPCAT (Treatment of Preserved Cardiac Function Heart Failure With an Aldosterone Antagonist) trial166 investigated the effects of spironolactone on a combined endpoint of death, aborted cardiac death, and HF hospitalization in patients with HFpEF. A small reduction (HR=0.89) in this composite endpoint did not reach statistical significance, although HF hospitalization was reduced (HR=0.83); known side effects of hyperkalemia and rising creatinine were seen more commonly in the treatment group.166 An unusual amount of regional variation was seen in this trial, prompting a post-hoc analysis167 that showed that rates of the primary endpoint were 4-fold lower in Russia/Georgia than in North America and South America (the Americas). Rates in the Americas were comparable to those in other HFpEF trials.169,170 The post-hoc analysis showed efficacy in the Americas (HR=0.83) but not in Russia/Georgia (HR=1.10). Moreover, a sample of the Russia/Georgia population, despite having been in the active treatment arm, had nondetectable levels of the metabolite of spironolactone. These post- hoc analyses have significant limitations, but they suggest that in appropriately selected patients with symptomatic HFpEF (with ejection fraction [EF] ≥45%, elevated BNP level or HF admission within 1 year, estimated glomerular filtration rate >30 mL/min creatinine <2.5 mg/dL, and potassium <5.0 mEq/L), particularly in those with elevated BNP levels, use of spironolactone might be considered with close monitoring of potassium and renal function. Confirmatory studies are required. With regard to the use of mineralocorticoid receptor antagonists, creatinine should be <2.5 mg/dL in men or <2.0 mg/dL in women (or estimated glomerular filtration rate >30 mL/min) and potassium should be <5.0 mEq/L. Careful monitoring of potassium, renal function, and diuretic dosing represents best practices at initiation and during follow-up thereafter to minimize risk of hyperkalemia and worsening renal function. The use of ARBs might be considered to decrease hospitalizations for 2013 recommendation remains current. IIb B patients with HFpEF. 169 III: No Routine use of nitrates or phosphodiesterase-5 inhibitors to increase NEW: Current recommendation reflects new B-R Benefit activity or QoL in patients with HFpEF is ineffective.171,172 data from RCTs. See Online Data Supplement C. Nitrate therapy can reduce pulmonary congestion and improve exercise tolerance in patients with HFrEF. However, the NEAT-HFpEF (Nitrate’s Effect on Activity Tolerance in Heart Failure With Preserved Ejection Fraction) trial171 randomized 110 patients with EF ≥50% on stable HF therapy, not including nitrates, and with activity limited by dyspnea, fatigue, or chest pain, to either isosorbide mononitrate or placebo and found no beneficial effects on activity levels, QoL, exercise tolerance, or NT-proBNP levels. On the basis of this trial, routine use of nitrates in patients with HFpEF is not recommended. This recommendation does not apply Downloaded from http://ahajournals.org by on August 12, 2021 to patients with HFpEF and symptomatic CAD for whom nitrates may provide symptomatic relief. Phosphodiesterase-5 inhibition augments the nitric oxide system by upregulating cGMP activity. The RELAX (Phosphodiesterase-5 Inhibition to Improve Clinical Status and Exercise Capacity in Heart Failure with Preserved Ejection 172 o Fraction) trial randomized 216 patients with EF ≥50% on stable HF therapy and with reduced exercise tolerance (peak observed V 2 <60% of predicted) to phosphodiesterase-5 inhibition with sildenafil or placebo. This study did not show improvement in oxygen consumption or exercise tolerance. III: No Routine use of nutritional supplements is not recommended for patients 2013 recommendation remains current. C Benefit with HFpEF.

9. IMPORTANT COMORBIDITIES IN HF 9.2. Anemia: Recommendations Recommendations for Anemia COR LOE Recommendations Comment/Rationale IIb B-R In patients with NYHA class II and III HF and iron deficiency (ferritin <100 NEW: New evidence consistent with ng/mL or 100 to 300 ng/mL if transferrin saturation is <20%), therapeutic benefit. See Online Data intravenous iron replacement might be reasonable to improve functional Supplement D. status and QoL.173,174 Routine baseline assessment of all patients with HF includes an evaluation for anemia in addition to other baseline laboratory measurements. Anemia is independently associated with HF disease severity, and iron deficiency appears to be uniquely associated with reduced exercise capacity. When iron deficiency is diagnosed and after full evaluation for cause, intravenous repletion of iron, especially in the setting of concomitant hepcidin deficiency in HF, may improve exercise capacity and QoL. Studies examining correction of iron deficiency in HF have demonstrated improvement in surrogate endpoints, such as QoL, NT- proBNP, and LVEF; however, controlled trials have been underpowered to detect reductions in hard clinical endpoints. The FAIR-HF (Ferric Carboxymaltose Assessment in Patients With Iron Deficiency and Chronic Heart Failure) trial173 demonstrated improvements in NYHA class and functional capacity over a short- term exposure. The CONFIRM-HF (Ferric Carboxymaltose Evaluation on Performance in Patients With Iron Deficiency in Combination with Chronic Heart Failure) trial174 included a larger cohort of patients (n=304) and demonstrated improvements in 6-minute walk test. A meta-analysis of 5 prospective controlled studies (631 patients) evaluated the effect of intravenous iron on deaths, hospitalizations, and other events in patients with HF and iron deficiency.175 Patients receiving intravenous iron experienced limited but statistically significant improvements in functional capacity and LVEF but no reduction in mortality rate. The FAIR-HF 2 trial is underway to further address the potential benefit of intravenous iron in HF associated with iron deficiency. Therefore, a strong recommendation for intravenous iron repletion must await the results of an appropriately powered trial on morbidity and mortality. There is an uncertain evidence base for oral iron repletion in the setting of anemia associated with HF.

Circulation. 2017;136:e137–e161. DOI: 10.1161/CIR.0000000000000509 August 8, 2017 e147 Yancy et al

Recommendations for Anemia (Continued) COR LOE Recommendations Comment/Rationale III: No In patients with HF and anemia, erythropoietin-stimulating agents NEW: Current recommendation reflects new B-R Benefit should not be used to improve morbidity and mortality.176 evidence demonstrating absence of therapeutic benefit. See Online Data Supplement D. Small studies evaluating the treatment of anemia in patients with HF have suggested a trend toward improvement in functional capacity and reduction in hospitalization with the use of erythropoietin-stimulating agents,177–182 but results have varied183 and have been limited because of sample size. Although a meta-analysis of 11 RCTs (n=794) comparing erythropoietin-stimulating agents to control in patients with HF demonstrated significant improvements in o 184 6-minute walk, exercise duration, peak V 2, NYHA functional status, EF, BNP, HF-related hospitalizations, and QoL, in the STAMINA-HeFT (Study of Anemia in Heart Failure) trial,183 darbepoetin alfa was not associated with significant clinical benefits. In the largest RCT to date (n=2278), correction of anemia with darbopoetin alfa did not result in benefit and resulted in a significant increase in the risk of thromboembolic events and a nonsignificant increase in fatal and nonfatal strokes, supporting findings from other trials.176,185–188 In summary, the strongest evidence on erythropoietin-stimulating agent therapy in HF suggests lack of benefit and increased adverse events. Therefore, erythropoietin-stimulating agent therapy cannot be recommended in patients with HF and anemia.

Table 3. Drugs Commonly Used for HFrEF (Stage C HF)

Mean Doses Achieved Drug Initial Daily Dose(s) Maximum Doses(s) in Clinical Trials References ACE inhibitors Captopril 6.25 mg TID 50 mg TID 122.7 mg QD 158 Enalapril 2.5 mg BID 10–20 mg BID 16.6 mg QD 129 Fosinopril 5–10 mg QD 40 mg QD N/A — Lisinopril 2.5–5 mg QD 20–40 mg QD 32.5–35.0 mg QD 130 Perindopril 2 mg QD 8–16 mg QD N/A — Quinapril 5 mg BID 20 mg BID N/A — Ramipril 1.25–2.5 mg QD 10 mg QD N/A — Trandolapril 1 mg QD 4 mg QD N/A — Downloaded from http://ahajournals.org by on August 12, 2021 ARBs Candesartan 4–8 mg QD 32 mg QD 24 mg QD 137 Losartan 25–50 mg QD 50–150 mg QD 129 mg QD 136 Valsartan 20–40 mg BID 160 mg BID 254 mg QD 134 ARNI Sacubitril/valsartan 49/51 mg BID (sacubitril/valsartan) (therapy 97/103 mg BID 375 mg QD; target dose: 24/26 mg, 138 may be initiated at 24/26 mg BID) (sacubitril/valsartan) 49/51 mg OR 97/103 mg BID

If channel inhibitor Ivabradine 5 mg BID 7.5 mg BID 6.4 mg BID (at 28 d) 6.5 mg 155–157 BID (at 1 y) Aldosterone antagonists Spironolactone 12.5–25 mg QD 25 mg QD or BID 26 mg QD 142 Eplerenone 25 mg QD 50 mg QD 42.6 mg QD 159 Beta blockers Bisoprolol 1.25 mg QD 10 mg QD 8.6 mg QD 160 Carvedilol 3.125 mg BID 50 mg BID 37 mg QD 161 Carvedilol CR 10 mg QD 80 mg QD N/A — Metoprolol succinate extended 12.5–25 mg QD 200 mg QD 159 mg QD 139 release (metoprolol CR/XL) Isosorbide dinitrate and hydralazine Fixed-dose combination 20 mg isosorbide dinitrate/37.5 mg 40 mg isosorbide dinitrate/75 90 mg isosorbide dinitrate/ 162 hydralazine TID mg hydralazine TID ~175 mg hydralazine QD Isosorbide dinitrate 20–30 mg isosorbide dinitrate/25–50 40 mg isosorbide dinitrate TID N/A 163 and hydralazine mg hydralazine TID or QD with 100 mg hydralazine TID

Modified (Table 15) from the 2013 HF guideline.9 ACE indicates angiotensin-converting enzyme; ARB, angiotensin receptor blocker; ARNI, angiotensin receptor-neprilysin inhibitor; BID, twice daily; CR, controlled release; CR/XL, controlled release/extended release; HF, heart failure; HFrEF, heart failure with reduced ejection fraction; N/A, not applicable; QD, once daily; and TID, 3 times daily.

e148 August 8, 2017 Circulation. 2017;136:e137–e161. DOI: 10.1161/CIR.0000000000000509 2017 ACC/AHA/HFSA Heart Failure Focused Update CLINICAL STATEMENTS AND GUIDELINES 9.5. Hypertension (New Section) 9.5.1. Treating Hypertension to Reduce the Incidence of HF: Recommendation Recommendation for Prevention COR LOE Recommendations Comment/Rationale I B-R In patients at increased risk, stage A HF, the optimal blood pressure in NEW: Recommendation reflects new RCT data. 189–193

those with hypertension should be less than 130/80 mm Hg. See Online Data Supplements E and F. A large RCT demonstrated that in those with increased cardiovascular risk (defined as age >75 years, established vascular disease, chronic renal disease, or a Framingham Risk Score >15%), control of blood pressure to a goal systolic pressure of <120 mm Hg, as determined by blood pressure assessment as per research protocol, was associated with a significant reduction in the incidence of HF191 and an overall decrease in cardiovascular death. Blood pressure measurements as generally taken in the office setting are typically 5 to 10 mm Hg higher than research measurements; thus, the goal of <130/80 mm Hg is an approximation of the target blood pressure in conventional practice. Targeting a significant reduction in systolic blood pressure in those at increased risk for cardiovascular disease is a novel strategy to prevent HF.

9.5.2. Treating Hypertension in Stage C HFrEF: Recommendation Recommendation for Hypertension in Stage C HFrEF COR LOE Recommendation Comment/Rationale I C-EO Patients with HFrEF and hypertension should be prescribed GDMT NEW: Recommendation has been adapted titrated to attain systolic blood pressure less than 130 mm Hg.191 from recent clinical trial data but not See Online Data specifically tested per se in a randomized trial Supplements E and F. of patients with HF. Clinical trials evaluating goal blood pressure reduction and optimal blood pressure–lowering agents in the setting of HFrEF and concomitant hypertension have not been done. However, it is apparent that in those patients at higher risk, blood pressure lowering is associated with fewer adverse cardiovascular events. GDMT for HFrEF with agents known to lower blood pressure should consider a goal blood pressure reduction consistent with a threshold now associated with improved clinical outcomes but not yet proven by RCTs in a population with HF.

9.5.3. Treating Hypertension in Stage C HFpEF: Recommendation Recommendation for Hypertension in Stage C HFpEF

Downloaded from http://ahajournals.org by on August 12, 2021 COR LOE Recommendation Comment/Rationale I C-LD Patients with HFpEF and persistent hypertension after management of NEW: New target goal blood pressure based volume overload should be prescribed GDMT titrated to attain systolic on updated interpretation of recent clinical See Online Data blood pressure less than 130 mm Hg.9,167,169,170,195–199 trial data. Supplements E and F. The use of nitrates in the setting of HFpEF is associated with a signal of harm and, in most situations, should be avoided. For many common antihypertensive agents, including alpha blockers, beta blockers, and calcium channel blockers, there are limited data to guide the choice of antihypertensive therapy in the setting of HFpEF. 172 Nevertheless, RAAS inhibition with ACE inhibitor, ARB (especially mineralocorticoid receptor antagonists), and possibly ARNI would represent the preferred choice. A shared decision-making discussion with the patient influenced by physician judgment should drive the ultimate choice of antihypertensive agents.

9.6. Sleep-Disordered Breathing: Recommendations (Moved from Section 7.3.1.4, Treatment of Sleep Disorders in the 2013 HF guideline.) Recommendations for Treatment of Sleep Disorders COR LOE Recommendations Comment/Rationale IIa C-LD In patients with NYHA class II–IV HF and suspicion of sleep-disordered NEW: Recommendation reflects clinical breathing or excessive daytime sleepiness, a formal sleep assessment is necessity to distinguish obstructive versus See Online Data reasonable.200,201 central sleep apnea. Supplement G. Sleep disorders are common in patients with HF. A study of adults with chronic HF treated with evidence-based therapies found that 61% had either central or obstructive sleep apnea.202 It is clinically important to distinguish obstructive sleep apnea from central sleep apnea, given the different responses to treatment. Adaptive servo-ventilation for central sleep apnea is associated with harm.203 Continuous positive airway pressure (CPAP) for obstructive sleep apnea improves sleep quality, reduces the apnea-hypopnea index, and improves nocturnal oxygenation.200,201 IIb B-R In patients with cardiovascular disease and obstructive sleep apnea, NEW: New data demonstrate the limited scope CPAP may be reasonable to improve sleep quality and daytime of benefit expected from CPAP for obstructive See Online Data sleepiness.204 sleep apnea. Supplement G. In patients with sleep apnea, a trial evaluated the impact of CPAP with usual therapy versus usual therapy alone on subsequent cardiovascular events, including HF. 204 In this RCT of >2700 patients, there was no evidence of benefit on cardiovascular events at a mean follow-up of 3.7 years for CPAP plus usual care compared with usual care alone. Improvements in sleep quality were noteworthy and represented the primary indication for initiating CPAP treatment.204 However, in patients with atrial fibrillation (AF) (a frequent comorbidity noted with HF), the use of CPAP for obstructive sleep apnea was helpful. In a trial of 10 132 patients with AF and obstructive sleep apnea, patients on CPAP treatment were less likely to progress to more permanent forms of AF than were patients without CPAP.205

Circulation. 2017;136:e137–e161. DOI: 10.1161/CIR.0000000000000509 August 8, 2017 e149 Yancy et al

Recommendations for Treatment of Sleep Disorders (Continued) COR LOE Recommendations Comment/Rationale III: Harm B-R In patients with NYHA class II–IV HFrEF and central sleep apnea, adaptive NEW: New data demonstrate a signal of harm servo-ventilation causes harm.203 when adaptive servo-ventilation is used for See Online Data central sleep apnea. Supplement G. Mortality rate (all cause and cardiovascular) was higher with adaptive servo-ventilation plus GDMT than with GDMT alone in a single RCT to test the addition of adaptive servo-ventilation (≥5 hours/night, 7 days/week) to GDMT in patients with HFrEF and central sleep apnea.203 A similar risk has been seen in another trial, and a third trial of adaptive servo-ventilation in central sleep apnea and HF was aborted because of ethical concerns. The weight of evidence does not support the use of adaptive servo-ventilation for central sleep apnea in HFrEF.

Gayle R. Whitman, PhD, RN, FAHA, FAAN, Senior Vice Presi- ACC/AHA TASK FORCE MEMBERS dent, Office of Science Operations Glenn N. Levine, MD, FACC, FAHA, Chair; Patrick T. O’Gara, Jody Hundley, Production Manager, Scientific Publications, MD, FACC, FAHA, Chair-Elect; Jonathan L. Halperin, MD, Office of Science Operations FACC, FAHA, Immediate Past Chair*; Sana M. Al-Khatib, MD, MHS, FACC, FAHA; Kim K. Birtcher, PharmD, MS, AACC; Bi- ykem Bozkurt, MD, PhD, FACC, FAHA; Ralph G. Brindis, MD, FOOTNOTES MPH, MACC*; Joaquin E. Cigarroa, MD, FACC; Lesley H. Cur- This document was approved by the American College of tis, PhD, FAHA; Lee A. Fleisher, MD, FACC, FAHA; Federico Gentile, MD, FACC; Samuel Gidding, MD, FAHA; Mark A. Cardiology Clinical Policy Approval Committee, the American Hlatky, MD, FACC; John Ikonomidis, MD, PhD, FAHA; José Heart Association Science Advisory and Coordinating Com- Joglar, MD, FACC, FAHA; Susan J. Pressler, PhD, RN, FAHA; mittee, the American Heart Association Executive Committee, Duminda N. Wijeysundera, MD, PhD and the Heart Failure Society of America Executive Committee in April 2017. The Comprehensive RWI Data Supplement table is avail- PRESIDENTS AND STAFF able with this article at http://circ.ahajournals.org/lookup/ suppl/doi:10.1161/CIR.0000000000000509/-/DC1. American College of Cardiology The Data Supplement is available with this article at

Downloaded from http://ahajournals.org by on August 12, 2021 Mary Norine Walsh, MD, FACC, President http://circ.ahajournals.org/lookup/suppl/doi:10.1161/ Shalom Jacobovitz, Chief Executive Officer CIR.0000000000000509/-/DC2. William J. Oetgen, MD, MBA, FACC, Executive Vice President, This article has been copublished in the Journal of the Science, Education, Quality, and Publishing American College of Cardiology and the Journal of Cardiac Amelia Scholtz, PhD, Publications Manager, Science, Education, Failure. Quality, and Publishing Copies: This document is available on the World Wide Web sites of the American College of Cardiology (www.acc.org), the American Heart Association (professional.heart.org), American College of Cardiology/ and the Heart Failure Society of America (www.hfsa.org). American Heart Association A copy of the document is available at http://professional. heart.org/statements by selecting either the “By Topic” Katherine Sheehan, PhD, Director of Guideline Strategy and link or the “By Publication Date” link. To purchase addi- Operations tional reprints, call 843-216-2533 or e-mail kelle.ramsay@ Lisa Bradfield, CAE, Director, Guideline Methodology and Policy wolterskluwer.com. Abdul R. Abdullah, MD, Science and Medicine Advisor Expert peer review of AHA Scientific Statements is con- Morgane Cibotti-Sun, MPH, Project Manager, Clinical Practice ducted by the AHA Office of Science Operations. For more Guidelines Sam Shahid, MBBS, MPH, Associate Science and Medicine on AHA statements and guidelines development, visit http:// Advisor professional.heart.org/statements and select the “Policies and Development” link. Permissions: Multiple copies, modification, alteration, American Heart Association enhancement, and/or distribution of this document are not Steven R. Houser, PhD, FAHA, President permitted without the express permission of the American Nancy Brown, Chief Executive Officer Heart Association. Instructions for obtaining permission are Rose Marie Robertson, MD, FAHA, Chief Science and Medicine located at http://www.heart.org/HEARTORG/General/Copyright- Officer Permission-Guidelines_UCM_300404_Article.jsp. A link to the “Copyright Permissions Request Form” appears on the *Former Task Force member; current member during the writing right side of the page. effort. Circulation is available at http://circ.ahajournals.org.

e150 August 8, 2017 Circulation. 2017;136:e137–e161. DOI: 10.1161/CIR.0000000000000509 2017 ACC/AHA/HFSA Heart Failure Focused Update CLINICAL STATEMENTS AND GUIDELINES suspected heart failure: results of the UK natriuretic peptide study. Eur J REFERENCES Heart Fail. 2005;7:537–41. 1. Committee on Standards for Developing Trustworthy Clinical Practice 18. Son CS, Kim YN, Kim HS, et al. Decision-making model for early diagnosis Guidelines, Institute of Medicine (US). Clinical Practice Guidelines We Can of congestive heart failure using rough set and decision tree approaches. Trust. Washington, DC: National Academies Press, 2011. J Biomed Inform. 2012;45:999–1008. 2. Committee on Standards for Systematic Reviews of Comparative Ef- 19. Kelder JC, Cramer MJ, Van WJ, et al. The diagnostic value of physical ex- fectiveness Research, Institute of Medicine (US). Finding What Works in amination and additional testing in primary care patients with suspected Health Care: Standards for Systematic Reviews. Washington, DC: National heart failure. Circulation. 2011;124:2865–73. Academies Press, 2011. 20. Balion C, Don-Wauchope A, Hill S, et al. Use of Natriuretic Peptide Mea-

3. Anderson JL, Heidenreich PA, Barnett PG, et al. ACC/AHA statement on surement in the Management of Heart Failure [Internet]. 13(14)-EHC118- cost/value methodology in clinical practice guidelines and performance EF ed. Rockville, MD: 2013. measures: a report of the American College of Cardiology/American Heart 21. Booth RA, Hill SA, Don-Wauchope A, et al. Performance of BNP and NT- Association Task Force on Performance Measures and Task Force on Prac- proBNP for diagnosis of heart failure in primary care patients: a systematic tice Guidelines. Circulation. 2014;129:2329–45. review. Heart Fail Rev. 2014;19:439–51. 4. ACCF/AHA Task Force on Practice Guidelines. Methodology Manual 22. Dao Q, Krishnaswamy P, Kazanegra R, et al. Utility of B-type natriuretic and Policies From the ACCF/AHA Task Force on Practice Guidelines; peptide in the diagnosis of congestive heart failure in an urgent-care set- 2010. American College of Cardiology and American Heart Association. ting. J Am Coll Cardiol. 2001;37:379–85. Available at: http://my.americanheart.org/idc/groups/ahamah-public/@ 23. Davis M, Espiner E, Richards G, et al. Plasma brain natriuretic peptide in wcm/@sop/documents/downloadable/ucm_319826.pdf Accessed May assessment of acute dyspnoea. Lancet. 1994;343:440–4. 5, 2017. 24. Maisel AS, Krishnaswamy P, Nowak RM, et al. Rapid measurement of B- 5. Halperin JL, Levine GN, Al-Khatib SM, et al. Further evolution of the type natriuretic peptide in the emergency diagnosis of heart failure. N Engl ACC/AHA clinical practice guideline recommendation classification sys- J Med. 2002;347:161–7. tem: a report of the American College of Cardiology/American Heart 25. Moe GW, Howlett J, Januzzi JL, et al. N-terminal pro-B-type natriuretic Association Task Force on Clinical Practice Guidelines. Circulation. peptide testing improves the management of patients with suspected 2016;133:1426–8. acute heart failure: primary results of the Canadian prospective random- 6. Jacobs AK, Kushner FG, Ettinger SM, et al. ACCF/AHA clinical practice ized multicenter IMPROVE-CHF study. Circulation. 2007;115:3103–10. guideline methodology summit report: a report of the American College 26. Mueller C, Scholer A, Laule-Kilian K, et al. Use of B-type natriuretic pep- of Cardiology Foundation/American Heart Association Task Force on Prac- tide in the evaluation and management of acute dyspnea. N Engl J Med. tice Guidelines. Circulation. 2013;127:268–310. 2004;350:647–54. 7. Jacobs AK, Anderson JL, Halperin JL. The evolution and future of ACC/ 27. van Kimmenade RR, Pinto YM, Bayes-Genis A, et al. Usefulness of in- AHA clinical practice guidelines: a 30-year journey: a report of the Ameri- termediate amino-terminal pro-brain natriuretic peptide concentra- can College of Cardiology/American Heart Association Task Force on Prac- tions for diagnosis and prognosis of acute heart failure. Am J Cardiol. tice Guidelines. Circulation. 2014;130:1208–17. 2006;98:386–90. 8. Arnett DK, Goodman RA, Halperin JL, et al. AHA/ACC/HHS strategies to 28. Januzzi JL Jr, Chen-Tournoux AA, Moe G. Amino-terminal pro-B-type na- enhance application of clinical practice guidelines in patients with cardio- triuretic peptide testing for the diagnosis or exclusion of heart failure in vascular disease and comorbid conditions: from the American Heart As- patients with acute symptoms. Am J Cardiol. 2008;101:29–38. sociation, American College of Cardiology, and US Department of Health 29. Santaguida PL, Don-Wauchope AC, Ali U, et al. Incremental value of Downloaded from http://ahajournals.org by on August 12, 2021 and Human Services. Circulation. 2014;130:1662–7. natriuretic peptide measurement in acute decompensated heart failure 9. Yancy CW, Jessup M, Bozkurt B, et al. 2013 ACCF/AHA guideline for the (ADHF): a systematic review. Heart Fail Rev. 2014;19:507–19. management of heart failure: a report of the American College of Cardiol- 30. Hill SA, Booth RA, Santaguida PL, et al. Use of BNP and NT-proBNP for ogy Foundation/American Heart Association Task Force on Practice Guide- the diagnosis of heart failure in the emergency department: a systematic lines. Circulation. 2013;128:e240–327. review of the evidence. Heart Fail Rev. 2014;19:421–38. 10. Yancy CW, Jessup M, Bozkurt B, et al. 2016 ACC/AHA/HFSA focused 31. Costello-Boerrigter LC, Boerrigter G, Redfield MM, et al. Amino-terminal update on new pharmacological therapy for heart failure: an update of pro-B-type natriuretic peptide and B-type natriuretic peptide in the gener- the 2013 ACCF/AHA guideline for the management of heart failure: a al community: determinants and detection of left ventricular dysfunction. report of the American College of Cardiology/American Heart Association J Am Coll Cardiol. 2006;47:345–53. Task Force on Clinical Practice Guidelines and the Heart Failure Society of 32. de Lemos JA, McGuire DK, Khera A, et al. Screening the population for America. Circulation. 2016;134:e282–93. left ventricular hypertrophy and left ventricular systolic dysfunction us- 11. Ponikowski P, Voors AA, Anker SD, et al. 2016 ESC Guidelines for the ing natriuretic peptides: results from the Dallas Heart Study. Am Heart J. diagnosis and treatment of acute and chronic heart failure: The Task Force 2009;157:746–53. for the diagnosis and treatment of acute and chronic heart failure of the 33. Goetze JP, Mogelvang R, Maage L, et al. Plasma pro-B-type natriuretic European Society of Cardiology (ESC). Developed with the special con- peptide in the general population: screening for left ventricular hypertro- tribution of the Heart Failure Association (HFA) of the ESC. Eur Heart J. phy and systolic dysfunction. Eur Heart J. 2006;27:3004–10. 2016;37:2129–200. 34. Ng LL, Loke IW, Davies JE, et al. Community screening for left ventricular 12. Packer M, McMurray JJ, Desai AS, et al. Angiotensin receptor neprilysin systolic dysfunction using plasma and urinary natriuretic peptides. J Am inhibition compared with enalapril on the risk of clinical progression in Coll Cardiol. 2005;45:1043–50. surviving patients with heart failure. Circulation. 2015;131:54–61. 35. Ho JE, Liu C, Lyass A, et al. Galectin-3, a marker of cardiac fibrosis, pre- 13. Zile MR, Claggett BL, Prescott MF, et al. Prognostic Implications of Changes dicts incident heart failure in the community. J Am Coll Cardiol. 2012;60: in N-Terminal Pro-B-Type Natriuretic Peptide in Patients With Heart Failure. 1249–56. J Am Coll Cardiol. 2016;68:2425–36. 36. Wang TJ, Wollert KC, Larson MG, et al. Prognostic utility of novel bio- 14. Solomon SD, Zile M, Pieske B, et al. The angiotensin receptor nepri- markers of cardiovascular stress: the Framingham Heart Study. Circulation. lysin inhibitor LCZ696 in heart failure with preserved ejection frac- 2012;126:1596–604. tion: a phase 2 double-blind randomised controlled trial. Lancet. 37. Xanthakis V, Larson MG, Wollert KC, et al. Association of novel biomark- 2012;380:1387–95. ers of cardiovascular stress with left ventricular hypertrophy and dysfunc- 15. Richards AM, Doughty R, Nicholls MG, et al. Plasma N-terminal pro-brain tion: implications for screening. J Am Heart Assoc. 2013;2:e000399. natriuretic peptide and adrenomedullin: prognostic utility and predic- 38. Anwaruddin S, Lloyd-Jones DM, Baggish A, et al. Renal function, con- tion of benefit from carvedilol in chronic ischemic left ventricular dys- gestive heart failure, and amino-terminal pro-brain natriuretic peptide function. Australia-New Zealand Heart Failure Group. J Am Coll Cardiol. measurement: results from the ProBNP Investigation of Dyspnea in the 2001;37:1781–7. Emergency Department (PRIDE) Study. J Am Coll Cardiol. 2006;47:91–7. 16. Tang WH, Girod JP, Lee MJ, et al. Plasma B-type natriuretic peptide lev- 39. Redfield MM, Rodeheffer RJ, Jacobsen SJ, et al. Plasma brain natriuretic els in ambulatory patients with established chronic symptomatic systolic peptide concentration: impact of age and gender. J Am Coll Cardiol. heart failure. Circulation. 2003;108:2964–6. 2002;40:976–82. 17. Zaphiriou A, Robb S, Murray-Thomas T, et al. The diagnostic accuracy of 40. Wang TJ, Larson MG, Levy D, et al. Impact of age and sex on plasma plasma BNP and NTproBNP in patients referred from primary care with natriuretic peptide levels in healthy adults. Am J Cardiol. 2002;90:254–8.

Circulation. 2017;136:e137–e161. DOI: 10.1161/CIR.0000000000000509 August 8, 2017 e151 Yancy et al

41. Chang AY, Abdullah SM, Jain T, et al. Associations among androgens, guidance for acute decompensated heart failure treatment by a pre- estrogens, and natriuretic peptides in young women: observations from defined NT-PRoBNP target on the reduction of readmIssion and Mortality the Dallas Heart Study. J Am Coll Cardiol. 2007;49:109–16. rAtes. Am Heart J. 2014;168:30–6. 42. Clerico A, Giannoni A, Vittorini S, et al. The paradox of low BNP levels in 62. Stienen S. PRIMA II: can NT-pro-brain-natriuretic peptide (NT-proBNP) obesity. Heart Fail Rev. 2012;17:81–96. guided therapy during admission for acute heart failure reduce mortality 43. De Vecchis R, Esposito C, Di Biase G, et al. B-type natriuretic peptide- and readmissions? J Card Fail. 2016;22:939–40. guided versus symptom-guided therapy in outpatients with chronic heart 63. Kociol RD, Pang PS, Gheorghiade M, et al. Troponin elevation in heart fail- failure: a systematic review with meta-analysis. J Cardiovasc Med (Hager- ure prevalence, mechanisms, and clinical implications. J Am Coll Cardiol. stown). 2014;15:122–34. 2010;56:1071–8. 44. Felker GM, Hasselblad V, Hernandez AF, et al. Biomarker-guided therapy 64. Amsterdam EA, Wenger NK, Brindis RG, et al. 2014 AHA/ACC guideline for the in chronic heart failure: a meta-analysis of randomized controlled trials. management of patients with non-ST-elevation acute coronary syndromes: Am Heart J. 2009;158:422–30. a report of the American College of Cardiology/American Heart Association 45. Li P, Luo Y, Chen YM. B-type natriuretic peptide-guided chronic heart Task Force on Practice Guidelines. Circulation. 2014;130:e344–426. failure therapy: a meta-analysis of 11 randomised controlled trials. Heart 65. Masson S, Anand I, Favero C, et al. Serial measurement of cardiac tro- Lung Circ. 2013;22:852–60. ponin T using a highly sensitive assay in patients with chronic heart fail- 46. Porapakkham P, Porapakkham P, Zimmet H, et al. B-type natriuretic ure: data from 2 large randomized clinical trials. Circulation. 2012;125: peptide-guided heart failure therapy: a meta-analysis. Arch Intern Med. 280–8. 2010;170:507–14. 66. Savarese G, Musella F, D'Amore C, et al. Changes of natriuretic peptides 47. Savarese G, Trimarco B, Dellegrottaglie S, et al. Natriuretic peptide-guided predict hospital admissions in patients with chronic heart failure: a meta- therapy in chronic heart failure: a meta-analysis of 2686 patients in 12 analysis. J Am Coll Cardiol HF. 2014;2:148–58. randomized trials. PLoS One. 2013;8:e58287. 67. de Lemos JA. Increasingly sensitive assays for cardiac troponins: a review. 48. Troughton RW, Frampton CM, Brunner-La Rocca HP, et al. Effect of B-type JAMA. 2013;309:2262–9. natriuretic peptide-guided treatment of chronic heart failure on total mor- 68. de Boer RA, Daniels LB, Maisel AS, et al. State of the Art: Newer biomark- tality and hospitalization: an individual patient meta-analysis. Eur Heart J. ers in heart failure. Eur J Heart Fail. 2015;17:559–69. 2014;35:1559–67. 69. Gopal DM, Sam F. New and emerging biomarkers in left ventricular sys- 49. Xin W, Lin Z, Mi S. Does B-type natriuretic peptide-guided therapy im- tolic dysfunction–insight into dilated cardiomyopathy. J Cardiovasc Transl prove outcomes in patients with chronic heart failure? A systematic re- Res. 2013;6:516–27. view and meta-analysis of randomized controlled trials. Heart Fail Rev. 70. O'Meara E, de DS, Rouleau JL, et al. Circulating biomarkers in patients 2015;20:69–80. with heart failure and preserved ejection fraction. Curr Heart Fail Rep. 50. Berger R, Moertl D, Peter S, et al. N-terminal pro-B-type natriuretic pep- 2013;10:350–8. tide-guided, intensive patient management in addition to multidisciplinary 71. Karayannis G, Triposkiadis F, Skoularigis J, et al. The emerging role of Ga- care in chronic heart failure: a 3-arm, prospective, randomized pilot study. lectin-3 and ST2 in heart failure: practical considerations and pitfalls using J Am Coll Cardiol. 2010;55:645–53. novel biomarkers. Curr Heart Fail Rep. 2013;10:441–9. 51. Eurlings LW, van Pol PE, Kok WE, et al. Management of chronic heart fail- 72. Ahmad T, Fiuzat M, Neely B, et al. Biomarkers of myocardial stress and ure guided by individual N-terminal pro-B-type natriuretic peptide targets: fibrosis as predictors of mode of death in patients with chronic heart fail- results of the PRIMA (Can PRo-brain-natriuretic peptide guided therapy ure. J Am Coll Cardiol HF. 2014;2:260–8. of chronic heart failure IMprove heart fAilure morbidity and mortality?) 73. Bayes-Genis A, de AM, Vila J, et al. Head-to-head comparison of 2 myo- Downloaded from http://ahajournals.org by on August 12, 2021 study. J Am Coll Cardiol. 2010;56:2090–100. cardial fibrosis biomarkers for long-term heart failure risk stratification: 52. Gaggin HK, Mohammed AA, Bhardwaj A, et al. Heart failure outcomes ST2 versus galectin-3. J Am Coll Cardiol. 2014;63:158–66. and benefits of NT-proBNP-guided management in the elderly: results 74. Gaggin HK, Szymonifka J, Bhardwaj A, et al. Head-to-head comparison of from the prospective, randomized ProBNP outpatient tailored chronic serial soluble ST2, growth differentiation factor-15, and highly-sensitive heart failure therapy (PROTECT) study. J Card Fail. 2012;18:626–34. troponin T measurements in patients with chronic heart failure. J Am Coll 53. Jourdain P, Jondeau G, Funck F, et al. Plasma brain natriuretic peptide- Cardiol HF. 2014;2:65–72. guided therapy to improve outcome in heart failure: the STARS-BNP Mul- 75. Ky B, French B, Levy WC, et al. Multiple biomarkers for risk prediction in ticenter Study. J Am Coll Cardiol. 2007;49:1733–9. chronic heart failure. Circ Heart Fail. 2012;5:183–90. 54. Karlstrom P, Alehagen U, Boman K, et al. Brain natriuretic peptide-guided 76. Sabatine MS, Morrow DA, de Lemos JA, et al. Evaluation of mul- treatment does not improve morbidity and mortality in extensively treated tiple biomarkers of cardiovascular stress for risk prediction and guiding patients with chronic heart failure: responders to treatment have a signifi- medical therapy in patients with stable coronary disease. Circulation. cantly better outcome. Eur J Heart Fail. 2011;13:1096–103. 2012;125:233–40. 55. Lainchbury JG, Troughton RW, Strangman KM, et al. N-terminal pro-B- 77. Ahmad T, Fiuzat M, Pencina MJ, et al. Charting a roadmap for heart failure type natriuretic peptide-guided treatment for chronic heart failure: results biomarker studies. J Am Coll Cardiol HF. 2014;2:477–88. from the BATTLESCARRED (NT-proBNP-Assisted Treatment To Lessen Serial 78. Miller WL, Hartman KA, Grill DE, et al. Serial measurements of midre- Cardiac Readmissions and Death) trial. J Am Coll Cardiol. 2009;55:53–60. gion proANP and copeptin in ambulatory patients with heart failure: in- 56. Persson H, Erntell H, Eriksson B, et al. Improved pharmacological therapy cremental prognostic value of novel biomarkers in heart failure. Heart. of chronic heart failure in primary care: a randomized Study of NT-proBNP 2012;98:389–94. Guided Management of Heart Failure–SIGNAL-HF (Swedish Intervention 79. Creemers EE, Tijsen AJ, Pinto YM. Circulating microRNAs: novel biomark- study–Guidelines and NT-proBNP AnaLysis in Heart Failure). Eur J Heart ers and extracellular communicators in cardiovascular disease? Circ Res. Fail. 2010;12:1300–8. 2012;110:483–95. 57. Pfisterer M, Buser P, Rickli H, et al. BNP-guided vs symptom-guided heart 80. Wong LL, Armugam A, Sepramaniam S, et al. Circulating microRNAs in failure therapy: the Trial of Intensified vs Standard Medical Therapy in El- heart failure with reduced and preserved left ventricular ejection fraction. derly Patients With Congestive Heart Failure (TIME-CHF) randomized trial. Eur J Heart Fail. 2015;17:393–404. JAMA. 2009;301:383–92. 81. Ovchinnikova ES, Schmitter D, Vegter EL, et al. Signature of circulat- 58. Shah MR, Califf RM, Nohria A, et al. The STARBRITE trial: a randomized, ing microRNAs in patients with acute heart failure. Eur J Heart Fail. pilot study of B-type natriuretic peptide-guided therapy in patients with 2016;18:414–23. advanced heart failure. J Card Fail. 2011;17:613–21. 82. Shah SH, Kraus WE, Newgard CB. Metabolomic profiling for the identifi- 59. Troughton RW, Frampton CM, Yandle TG, et al. Treatment of heart failure cation of novel biomarkers and mechanisms related to common cardio- guided by plasma aminoterminal brain natriuretic peptide (N-BNP) con- vascular diseases: form and function. Circulation. 2012;126:1110–20. centrations. Lancet. 2000;355:1126–30. 83. Cheng ML, Wang CH, Shiao MS, et al. Metabolic disturbances identified 60. Singer AJ, Birkhahn RH, Guss D, et al. Rapid Emergency Department Heart in plasma are associated with outcomes in patients with heart failure: Failure Outpatients Trial (REDHOT II): a randomized controlled trial of the diagnostic and prognostic value of metabolomics. J Am Coll Cardiol. effect of serial B-type natriuretic peptide testing on patient management. 2015;65:1509–20. Circ Heart Fail. 2009;2:287–93. 84. Zheng Y, Yu B, Alexander D, et al. Associations between metabolomic 61. Stienen S, Salah K, Moons AH, et al. Rationale and design of PRIMA II: compounds and incident heart failure among African Americans: the A multicenter, randomized clinical trial to study the impact of in-hospital ARIC Study. Am J Epidemiol. 2013;178:534–42.

e152 August 8, 2017 Circulation. 2017;136:e137–e161. DOI: 10.1161/CIR.0000000000000509 2017 ACC/AHA/HFSA Heart Failure Focused Update CLINICAL STATEMENTS AND GUIDELINES 85. Ledwidge M, Gallagher J, Conlon C, et al. Natriuretic peptide-based 108. Salah K, Kok WE, Eurlings LW, et al. A novel discharge risk model for screening and collaborative care for heart failure: the STOP-HF random- patients hospitalised for acute decompensated heart failure incorporat- ized trial. JAMA. 2013;310:66–74. ing N-terminal pro-B-type natriuretic peptide levels: a European coLlabo- 86. Huelsmann M, Neuhold S, Resl M, et al. PONTIAC (NT-proBNP selected ration on Acute decompeNsated Heart Failure: ELAN-HF Score. Heart. prevention of cardiac events in a population of diabetic patients without 2014;100:115–25. a history of cardiac disease): a prospective randomized controlled trial. 109. Flint KM, Allen LA, Pham M, et al. B-type natriuretic peptide predicts 30- J Am Coll Cardiol. 2013;62:1365–72. day readmission for heart failure but not readmission for other causes. J 87. Anand IS, Fisher LD, Chiang YT, et al. Changes in brain natriuretic pep- Am Heart Assoc. 2014;3:e000806. tide and norepinephrine over time and mortality and morbidity in the 110. Kociol RD, Horton JR, Fonarow GC, et al. Admission, discharge, or

Valsartan Heart Failure Trial (Val-HeFT). Circulation. 2003;107:1278–83. change in B-type natriuretic peptide and long-term outcomes: data 88. Berger R, Huelsman M, Strecker K, et al. B-type natriuretic peptide pre- from Organized Program to Initiate Lifesaving Treatment in Hospitalized dicts sudden death in patients with chronic heart failure. Circulation. Patients with Heart Failure (OPTIMIZE-HF) linked to Medicare claims. Circ 2002;105:2392–7. Heart Fail. 2011;4:628–36. 89. Forfia PR, Watkins SP, Rame JE, et al. Relationship between B-type natri- 111. Kociol RD, McNulty SE, Hernandez AF, et al. Markers of decongestion, uretic peptides and pulmonary capillary wedge pressure in the intensive dyspnea relief, and clinical outcomes among patients hospitalized with care unit. J Am Coll Cardiol. 2005;45:1667–71. acute heart failure. Circ Heart Fail. 2013;6:240–5. 90. Maeda K, Tsutamoto T, Wada A, et al. High levels of plasma brain natri- 112. Verdiani V, Ognibene A, Rutili MS, et al. NT-ProBNP reduction percent- uretic peptide and interleukin-6 after optimized treatment for heart fail- age during hospital stay predicts long-term mortality and readmission in ure are independent risk factors for morbidity and mortality in patients heart failure patients. J Cardiovasc Med (Hagerstown). 2008;9:694–9. with congestive heart failure. J Am Coll Cardiol. 2000;36:1587–93. 113. Bayes-Genis A, Lopez L, Zapico E, et al. NT-ProBNP reduction percentage 91. Neuhold S, Huelsmann M, Strunk G, et al. Comparison of copeptin, during admission for acutely decompensated heart failure predicts long- B-type natriuretic peptide, and amino-terminal pro-B-type natriuretic term cardiovascular mortality. J Card Fail. 2005;11:S3–8. peptide in patients with chronic heart failure: prediction of death at dif- 114. Alonso-Martinez JL, Llorente-Diez B, Echegaray-Agara M, et al. C-reactive ferent stages of the disease. J Am Coll Cardiol. 2008;52:266–72. protein as a predictor of improvement and readmission in heart failure. 92. Taub PR, Daniels LB, Maisel AS. Usefulness of B-type natriuretic peptide Eur J Heart Fail. 2002;4:331–6. levels in predicting hemodynamic and clinical decompensation. Heart Fail 115. Dieplinger B, Gegenhuber A, Kaar G, et al. Prognostic value of estab- Clin. 2009;5:169–75. lished and novel biomarkers in patients with shortness of breath attend- 93. Bettencourt P, Azevedo A, Pimenta J, et al. N-terminal-pro-brain natri- ing an emergency department. Clin Biochem. 2010;43:714–9. uretic peptide predicts outcome after hospital discharge in heart failure 116. Januzzi JL Jr, Peacock WF, Maisel AS, et al. Measurement of the inter- patients. Circulation. 2004;110:2168–74. leukin family member ST2 in patients with acute dyspnea: results from 94. Cheng V, Kazanagra R, Garcia A, et al. A rapid bedside test for B-type the PRIDE (Pro-Brain Natriuretic Peptide Investigation of Dyspnea in the peptide predicts treatment outcomes in patients admitted for decom- Emergency Department) study. J Am Coll Cardiol. 2007;50:607–13. pensated heart failure: a pilot study. J Am Coll Cardiol. 2001;37:386–91. 117. Manzano-Fernandez S, Mueller T, Pascual-Figal D, et al. Usefulness of 95. Fonarow GC, Peacock WF, Horwich TB, et al. Usefulness of B-type natri- soluble concentrations of interleukin family member ST2 as predictor of uretic peptide and cardiac troponin levels to predict in-hospital mortality mortality in patients with acutely decompensated heart failure relative to from ADHERE. Am J Cardiol. 2008;101:231–7. left ventricular ejection fraction. Am J Cardiol. 2011;107:259–67. 96. Logeart D, Thabut G, Jourdain P, et al. Predischarge B-type natriuretic 118. Rehman SU, Mueller T, Januzzi JL Jr. Characteristics of the novel interleu- Downloaded from http://ahajournals.org by on August 12, 2021 peptide assay for identifying patients at high risk of re-admission after kin family biomarker ST2 in patients with acute heart failure. J Am Coll decompensated heart failure. J Am Coll Cardiol. 2004;43:635–41. Cardiol. 2008;52:1458–65. 97. Maisel A, Hollander JE, Guss D, et al. Primary results of the Rapid 119. Shah RV, Chen-Tournoux AA, Picard MH, et al. Galectin-3, cardiac struc- Emergency Department Heart Failure Outpatient Trial (REDHOT). A multi- ture and function, and long-term mortality in patients with acutely de- center study of B-type natriuretic peptide levels, emergency department compensated heart failure. Eur J Heart Fail. 2010;12:826–32. decision making, and outcomes in patients presenting with shortness of 120. de Boer RA, Lok DJ, Jaarsma T, et al. Predictive value of plasma galectin-3 breath. J Am Coll Cardiol. 2004;44:1328–33. levels in heart failure with reduced and preserved ejection fraction. Ann 98. Zairis MN, Tsiaousis GZ, Georgilas AT, et al. Multimarker strategy for the Med. 2011;43:60–8. prediction of 31 days cardiac death in patients with acutely decompen- 121. Lok DJ, van der Meer P, de la Porte PW, et al. Prognostic value of galec- sated chronic heart failure. Int J Cardiol. 2010;141:284–90. tin-3, a novel marker of fibrosis, in patients with chronic heart failure: 99. Peacock WFI, De Marco T, Fonarow GC, et al. Cardiac troponin and out- data from the DEAL-HF study. Clin Res Cardiol. 2010;99:323–8. come in acute heart failure. N Engl J Med. 2008;358:2117–26. 122. Tang WH, Shrestha K, Shao Z, et al. Usefulness of plasma galectin-3 levels 100. Lee DS, Stitt A, Austin PC, et al. Prediction of heart failure mortality in in systolic heart failure to predict renal insufficiency and survival. Am J emergent care: a cohort study. Ann Intern Med. 2012;156:767–75. Cardiol. 2011;108:385–90. 101. Santaguida PL, Don-Wauchope AC, Oremus M, et al. BNP and NT-proBNP 123. Tang WH, Wu Y, Grodin JL, et al. Prognostic Value of Baseline and as prognostic markers in persons with acute decompensated heart fail- Changes in Circulating Soluble ST2 Levels and the Effects of Nesiritide in ure: a systematic review. Heart Fail Rev. 2014;19:453–70. Acute Decompensated Heart Failure. J Am Coll Cardiol HF. 2016;4:68–77. 102. Horwich TB, Patel J, MacLellan WR, et al. Cardiac troponin I is associ- 124. Januzzi JL, Mebazaa A, Di SS. ST2 and prognosis in acutely decompen- ated with impaired hemodynamics, progressive left ventricular dysfunc- sated heart failure: the International ST2 Consensus Panel. Am J Cardiol. tion, and increased mortality rates in advanced heart failure. Circulation. 2015;115:26B–31B. 2003;108:833–8. 125. Mebazaa A, Di SS, Maisel AS, et al. ST2 and multimarker testing in acute 103. Ilva T, Lassus J, Siirila-Waris K, et al. Clinical significance of cardiac tropo- decompensated heart failure. Am J Cardiol. 2015;115:38B–43B. nins I and T in acute heart failure. Eur J Heart Fail. 2008;10:772–9. 126. Fermann GJ, Lindsell CJ, Storrow AB, et al. Galectin 3 complements BNP 104. Dhaliwal AS, Deswal A, Pritchett A, et al. Reduction in BNP levels with in risk stratification in acute heart failure. Biomarkers. 2012;17:706–13. treatment of decompensated heart failure and future clinical events. 127. Lassus J, Gayat E, Mueller C, et al. Incremental value of biomarkers to J Card Fail. 2009;15:293–9. clinical variables for mortality prediction in acutely decompensated heart 105. O'Connor CM, Hasselblad V, Mehta RH, et al. Triage after hospitalization failure: the Multinational Observational Cohort on Acute Heart Failure with advanced heart failure: the ESCAPE (Evaluation Study of Congestive (MOCA) study. Int J Cardiol. 2013;168:2186–94. Heart Failure and Pulmonary Artery Catheterization Effectiveness) risk 128. Effects of enalapril on mortality in severe congestive heart failure. model and discharge score. J Am Coll Cardiol. 2010;55:872–8. Results of the Cooperative North Scandinavian Enalapril Survival Study 106. O'Brien RJ, Squire IB, Demme B, et al. Pre-discharge, but not admission, (CONSENSUS). The CONSENSUS Trial Study Group. N Engl J Med. levels of NT-proBNP predict adverse prognosis following acute LVF. Eur J 1987;316:1429–35. Heart Fail. 2003;5:499–506. 129. Effect of enalapril on survival in patients with reduced left ventricular 107. Cohen-Solal A, Logeart D, Huang B, et al. Lowered B-type natriuretic ejection fractions and congestive heart failure. The SOLVD Investigators. peptide in response to levosimendan or dobutamine treatment is associ- N Engl J Med. 1991;325:293–302. ated with improved survival in patients with severe acutely decompen- 130. Packer M, Poole-Wilson PA, Armstrong PW, et al. Comparative effects sated heart failure. J Am Coll Cardiol. 2009;53:2343–8. of low and high doses of the angiotensin-converting enzyme inhibitor,

Circulation. 2017;136:e137–e161. DOI: 10.1161/CIR.0000000000000509 August 8, 2017 e153 Yancy et al

lisinopril, on morbidity and mortality in chronic heart failure. ATLAS Study 153. Ruilope LM, Dukat A, Böhm M, et al. Blood-pressure reduction with Group. Circulation. 1999;100:2312–8. LCZ696, a novel dual-acting inhibitor of the angiotensin II receptor and 131. Pfeffer MA, Braunwald E, Moyé LA, et al. Effect of captopril on mortality neprilysin: a randomised, double-blind, placebo-controlled, active com- and morbidity in patients with left ventricular dysfunction after myocar- parator study. Lancet. 2010;375:1255–66. dial infarction: results of the Survival and Ventricular Enlargement Trial. 154. Bohm M, Robertson M, Ford I, et al. Influence of cardiovascular and non- The SAVE Investigators. N Engl J Med. 1992;327:669–77. cardiovascular co-morbidities on outcomes and treatment effect of heart 132. Effect of ramipril on mortality and morbidity of survivors of acute myocar- rate reduction with ivabradine in stable heart failure (from the SHIFT dial infarction with clinical evidence of heart failure. The Acute Infarction Trial). Am J Cardiol. 2015;116:1890–7. Ramipril Efficacy (AIRE) Study Investigators. Lancet. 1993;342:821–8. 155. Swedberg K, Komajda M, Böhm M, et al. Ivabradine and outcomes in 133. Kober L, Torp-Pedersen C, Carlsen JE, et al. A clinical trial of the an- chronic heart failure (SHIFT): a randomised placebo-controlled study. giotensin-converting-enzyme inhibitor trandolapril in patients with left Lancet. 2010;376:875–85. ventricular dysfunction after myocardial infarction. Trandolapril Cardiac 156. Fox K, Ford I, Steg PG, et al. Ivabradine in stable coronary artery disease Evaluation (TRACE) Study Group. N Engl J Med. 1995;333:1670–6. without clinical heart failure. N Engl J Med. 2014;371:1091–9. 134. Cohn JN, Tognoni G, Investigators VHFT. A randomized trial of the angio- 157. Fox K, Ford I, Steg PG, et al. Ivabradine for patients with stable coro- tensin-receptor blocker valsartan in chronic heart failure. N Engl J Med. nary artery disease and left-ventricular systolic dysfunction (BEAUTIFUL): 2001;345:1667–75. a randomised, double-blind, placebo-controlled trial. Lancet. 2008;372: 135. Pfeffer MA, McMurray JJV, Velazquez EJ, et al. Valsartan, captopril, or 807–16. both in myocardial infarction complicated by heart failure, left ventricular 158. Pitt B, Segal R, Martinez FA, et al. Randomised trial of losartan versus dysfunction, or both. N Engl J Med. 2003;349:1893–906. captopril in patients over 65 with heart failure (Evaluation of Losartan in 136. Konstam MA, Neaton JD, Dickstein K, et al. Effects of high-dose ver- the Elderly Study, ELITE). Lancet. 1997;349:747–52. sus low-dose losartan on clinical outcomes in patients with heart failure 159. Pitt B, Remme W, Zannad F, et al. Eplerenone, a selective aldosterone (HEAAL study): a randomised, double-blind trial. Lancet. 2009;374: blocker, in patients with left ventricular dysfunction after myocardial in- 1840–8. farction. N Engl J Med. 2003;348:1309–21. 137. Pfeffer MA, Swedberg K, Granger CB, et al. Effects of candesartan 160. Authors CI. The Cardiac Insufficiency Bisoprolol Study II (CIBIS-II): a ran- on mortality and morbidity in patients with chronic heart failure: the domised trial. Lancet. 1999;353:9–13. CHARM-Overall programme. Lancet. 2003;362:759–66. 161. Packer M, Coats AJ, Fowler MB, et al. Effect of carvedilol on survival in 138. McMurray JJV, Packer M, Desai AS, et al. Angiotensin-neprilysin inhibi- severe chronic heart failure. N Engl J Med. 2001;344:1651–8. tion versus enalapril in heart failure. N Engl J Med. 2014;371:993–1004. 162. Taylor AL, Ziesche S, Yancy C, et al. Combination of isosorbide di- 139. Effect of metoprolol CR/XL in chronic heart failure: Metoprolol CR/XL nitrate and hydralazine in blacks with heart failure. N Engl J Med. Randomised Intervention Trial in Congestive Heart Failure (MERIT-HF). 2004;351:2049–57. Lancet. 1999;353:2001–7. 163. Cohn JN, Archibald DG, Ziesche S, et al. Effect of vasodilator therapy 140. Packer M, Fowler MB, Roecker EB, et al. Effect of carvedilol on the on mortality in chronic congestive heart failure. Results of a Veterans morbidity of patients with severe chronic heart failure: results of the Administration Cooperative Study. N Engl J Med. 1986;314:1547–52. carvedilol prospective randomized cumulative survival (COPERNICUS) 164. Chobanian AV, Bakris GL, Black HR, et al. Seventh report of the Joint study. Circulation. 2002;106:2194–9. National Committee on Prevention, Detection, Evaluation, and Treatment 141. Eschalier R, McMurray JJV, Swedberg K, et al. Safety and efficacy of of High Blood Pressure. Hypertension. 2003;42:1206–52. eplerenone in patients at high risk for hyperkalemia and/or worsening re- 165. Levy D, Larson MG, Vasan RS, et al. The progression from hypertension Downloaded from http://ahajournals.org by on August 12, 2021 nal function: analyses of the EMPHASIS-HF study subgroups (Eplerenone to congestive heart failure. JAMA. 1996;275:1557–62. in Mild Patients Hospitalization And SurvIval Study in Heart Failure). J Am 166. Pitt B, Pfeffer MA, Assmann SF, et al. Spironolactone for heart failure with Coll Cardiol. 2013;62:1585–93. preserved ejection fraction. N Engl J Med. 2014;370:1383–92. 142. Pitt B, Zannad F, Remme WJ, et al. The effect of spironolactone 167. Pfeffer MA, Claggett B, Assmann SF, et al. Regional variation in patients on morbidity and mortality in patients with severe heart failure. and outcomes in the Treatment of Preserved Cardiac Function Heart Randomized Aldactone Evaluation Study Investigators. N Engl J Med. Failure With an Aldosterone Antagonist (TOPCAT) trial. Circulation. 1999;341:709–17. 2015;131:34–42. 143. Garg R, Yusuf S. Overview of randomized trials of angiotensin-converting 168. Edelmann F, Wachter R, Schmidt AG, et al. Effect of spironolactone on enzyme inhibitors on mortality and morbidity in patients with heart failure. diastolic function and exercise capacity in patients with heart failure with Collaborative Group on ACE Inhibitor Trials. JAMA. 1995;273:1450–6. preserved ejection fraction: the Aldo-DHF randomized controlled trial. 144. Woodard-Grice AV, Lucisano AC, Byrd JB, et al. Sex-dependent and JAMA. 2013;309:781–91. race-dependent association of XPNPEP2 C-2399A polymorphism with 169. Yusuf S, Pfeffer MA, Swedberg K, et al. Effects of candesartan in patients angiotensin-converting enzyme inhibitor-associated angioedema. with chronic heart failure and preserved left-ventricular ejection fraction: Pharmacogenet Genomics. 2010;20:532–6. the CHARM-Preserved Trial. Lancet. 2003;362:777–81. 145. Yusuf S, Teo KK, Pogue J, et al. Telmisartan, ramipril, or both in patients 170. Massie BM, Carson PE, McMurray JJ, et al. Irbesartan in patients at high risk for vascular events. N Engl J Med. 2008;358:1547–59. with heart failure and preserved ejection fraction. N Engl J Med. 146. Yusuf S, Teo K, Anderson C, et al. Effects of the angiotensin-receptor 2008;359:2456–67. blocker telmisartan on cardiovascular events in high-risk patients intol- 171. Redfield MM, Anstrom KJ, Levine JA, et al. Isosorbide Mononitrate in Heart erant to angiotensin-converting enzyme inhibitors: a randomised con- Failure with Preserved Ejection Fraction. N Engl J Med. 2015;373:2314–24. trolled trial. Lancet. 2008;372:1174–83. 172. Redfield MM, Chen HH, Borlaug BA, et al. Effect of phosphodiester- 147. Entresto [package insert]. Hanover, NJ: Novartis Pharmaceuticals ase-5 inhibition on exercise capacity and clinical status in heart failure Corporation, 2015. with preserved ejection fraction: a randomized clinical trial. JAMA. 148. Packer M, Califf RM, Konstam MA, et al. Comparison of omapatrilat and 2013;309:1268–77. enalapril in patients with chronic heart failure: the Omapatrilat Versus 173. Anker SD, Comin CJ, Filippatos G, et al. Ferric carboxymaltose in patients Enalapril Randomized Trial of Utility in Reducing Events (OVERTURE). with heart failure and iron deficiency. N Engl J Med. 2009;361:2436–48. Circulation. 2002;106:920–6. 174. Ponikowski P, van Veldhuisen DJ, Comin-Colet J, et al. Beneficial effects 149. Kostis JB, Packer M, Black HR, et al. Omapatrilat and enalapril in pa- of long-term intravenous iron therapy with ferric carboxymaltose in pa- tients with hypertension: the Omapatrilat Cardiovascular Treatment vs. tients with symptomatic heart failure and iron deficiency. Eur Heart J. Enalapril (OCTAVE) trial. Am J Hypertens. 2004;17:103–11. 2015;36:657–68. 150. Vardeny O, Miller R, Solomon SD. Combined neprilysin and renin- 175. Kapoor M, Schleinitz MD, Gemignani A, et al. Outcomes of patients angiotensin system inhibition for the treatment of heart failure. J Am with chronic heart failure and iron deficiency treated with intrave- Coll Cardiol HF. 2014;2:663–70. nous iron: a meta-analysis. Cardiovasc Hematol Disord. Drug Targets. 151. Messerli FH, Nussberger J. Vasopeptidase inhibition and angio-oedema. 2013;13:35–44. Lancet. 2000;356:608–9. 176. Swedberg K, Young JB, Anand IS, et al. Treatment of anemia with darbe- 152. Braunwald E. The path to an angiotensin receptor antagonist-neprilysin poetin alfa in systolic heart failure. N Engl J Med. 2013;368:1210–19. inhibitor in the treatment of heart failure. J Am Coll Cardiol. 177. Cleland JG, Sullivan JT, Ball S, et al. Once-monthly administration of dar- 2015;65:1029–41. bepoetin alfa for the treatment of patients with chronic heart failure

e154 August 8, 2017 Circulation. 2017;136:e137–e161. DOI: 10.1161/CIR.0000000000000509 2017 ACC/AHA/HFSA Heart Failure Focused Update CLINICAL STATEMENTS AND GUIDELINES and anemia: a pharmacokinetic and pharmacodynamic investigation. in adults aged >/=75 years: A randomized clinical trial. JAMA. J Cardiovasc Pharmacol. 2005;46:155–61. 2016;315:2673–82. 178. Klapholz M, Abraham WT, Ghali JK, et al. The safety and tolerability of 193. Lv J, Ehteshami P, Sarnak MJ, et al. Effects of intensive blood pressure darbepoetin alfa in patients with anaemia and symptomatic heart failure. lowering on the progression of chronic kidney disease: a systematic re- Eur J Heart Fail. 2009;11:1071–7. view and meta-analysis. CMAJ. 2013;185:949–57. 179. Ponikowski P, Anker SD, Szachniewicz J, et al. Effect of darbepoetin alfa 194. Deleted in press. on exercise tolerance in anemic patients with symptomatic chronic heart 195. Aronow WS, Fleg JL, Pepine CJ, et al. ACCF/AHA 2011 expert consen- failure: a randomized, double-blind, placebo-controlled trial. J Am Coll sus document on hypertension in the elderly: a report of the American Cardiol. 2007;49:753–62. College of Cardiology Foundation Task Force on Clinical Expert

180. Silverberg DS, Wexler D, Sheps D, et al. The effect of correction of mild Consensus Documents. Developed in collaboration with the American anemia in severe, resistant congestive heart failure using subcutaneous Academy of Neurology, American Geriatrics Society, American Society erythropoietin and intravenous iron: a randomized controlled study. J Am for Preventive Cardiology, American Society of Hypertension, American Coll Cardiol. 2001;37:1775–80. Society of Nephrology, Association of Black Cardiologists, and European 181. van der Meer P, Groenveld HF, Januzzi JL Jr, et al. Erythropoietin treat- Society of Hypertension. Circulation. 2011;123:2434–506. ment in patients with chronic heart failure: a meta-analysis. Heart. 196. Law MR, Morris JK, Wald NJ. Use of blood pressure lowering drugs in the 2009;95:1309–14. prevention of cardiovascular disease: meta-analysis of 147 randomised 182. van Veldhuisen DJ, Dickstein K, Cohen-Solal A, et al. Randomized, trials in the context of expectations from prospective epidemiological double-blind, placebo-controlled study to evaluate the effect of two studies. BMJ. 2009;338:b1665. dosing regimens of darbepoetin alfa in patients with heart failure and 197. Aronow WS, Ahn C, Kronzon I. Effect of propranolol versus no pro- anaemia. Eur Heart J. 2007;28:2208–16. pranolol on total mortality plus nonfatal myocardial infarction in older 183. Ghali JK, Anand IS, Abraham WT, et al. Randomized double-blind trial of patients with prior myocardial infarction, congestive heart failure, and darbepoetin alfa in patients with symptomatic heart failure and anemia. left ventricular ejection fraction > or = 40% treated with diuretics plus Circulation. 2008;117:526–35. angiotensin-converting enzyme inhibitors. Am J Cardiol. 1997;80:207–9. 184. Kotecha D, Ngo K, Walters JA, et al. Erythropoietin as a treatment of 198. van Veldhuisen DJ, Cohen-Solal A, Bohm M, et al. Beta-blockade with anemia in heart failure: systematic review of randomized trials. Am Heart nebivolol in elderly heart failure patients with impaired and preserved J. 2011;161:822–31. left ventricular ejection fraction: Data From SENIORS (Study of Effects 185. Bennett CL, Silver SM, Djulbegovic B, et al. Venous thromboembolism of Nebivolol Intervention on Outcomes and Rehospitalization in Seniors and mortality associated with recombinant erythropoietin and darbepoetin With Heart Failure). J Am Coll Cardiol. 2009;53:2150–58. administration for the treatment of cancer-associated anemia. JAMA. 199. Piller LB, Baraniuk S, Simpson LM, et al. Long-term follow-up of participants 2008;299:914–24. with heart failure in the antihypertensive and lipid-lowering treatment to 186. Bohlius J, Wilson J, Seidenfeld J, et al. Recombinant human erythropoi- prevent heart attack trial (ALLHAT). Circulation. 2011;124:1811–8. etins and cancer patients: updated meta-analysis of 57 studies including 200. Arzt M, Floras JS, Logan AG, et al. Suppression of central sleep apnea by 9353 patients. J Natl Cancer Inst. 2006;98:708–14. continuous positive airway pressure and transplant-free survival in heart 187. Pfeffer MA, Burdmann EA, Chen CY, et al. A trial of darbepoe- failure: a post hoc analysis of the Canadian Continuous Positive Airway tin alfa in type 2 diabetes and chronic kidney disease. N Engl J Med. Pressure for Patients with Central Sleep Apnea and Heart Failure Trial 2009;361:2019–32. (CANPAP). Circulation. 2007;115:3173–80. 188. Szczech LA, Barnhart HX, Inrig JK, et al. Secondary analysis of the CHOIR 201. Bradley TD, Logan AG, Kimoff RJ, et al. Continuous positive air- Downloaded from http://ahajournals.org by on August 12, 2021 trial epoetin-alpha dose and achieved hemoglobin outcomes. Kidney Int. way pressure for central sleep apnea and heart failure. N Engl J Med. 2008;74:791–8. 2005;353:2025–33. 189. Xie X, Atkins E, Lv J, et al. Effects of intensive blood pressure lowering 202. MacDonald M, Fang J, Pittman SD, et al. The current prevalence of sleep on cardiovascular and renal outcomes: updated systematic review and disordered breathing in congestive heart failure patients treated with meta-analysis. Lancet. 2016;387:435–43. beta-blockers. J Clin Sleep Med. 2008;4:38–42. 190. Thomopoulos C, Parati G, Zanchetti A. Effects of blood pressure lowering 203. Cowie MR, Woehrle H, Wegscheider K, et al. Adaptive Servo-Ventilation on outcome incidence in hypertension: 7. Effects of more vs. less inten- for Central Sleep Apnea in Systolic Heart Failure. N Engl J Med. sive blood pressure lowering and different achieved blood pressure levels 2015;373:1095–105. - updated overview and meta-analyses of randomized trials. J Hypertens. 204. McEvoy RD, Antic NA, Heeley E, et al. CPAP for prevention of cardiovas- 2016;34:613–22. cular events in obstructive sleep apnea. N Engl J Med. 2016;375:919–31. 191. Wright JT Jr, Williamson JD, Whelton PK, et al. A Randomized Trial 205. Holmqvist F, Guan N, Zhu Z, et al. Impact of obstructive sleep apnea of Intensive versus Standard Blood-Pressure Control. N Engl J Med. and continuous positive airway pressure therapy on outcomes in pa- 2015;373:2103–16. tients with atrial fibrillation-Results from the Outcomes Registry for 192. Williamson JD, Supiano MA, Applegate WB, et al. Intensive vs stan- Better Informed Treatment of Atrial Fibrillation (ORBIT-AF). Am Heart J. dard blood pressure control and cardiovascular disease outcomes 2015;169:647–54.

Circulation. 2017;136:e137–e161. DOI: 10.1161/CIR.0000000000000509 August 8, 2017 e155 Yancy et al

Appendix 1. Author Relationships With Industry and Other Entities (Relevant)—2017 ACC/AHA/HFSA Focused Update of the 2013 ACCF/AHA Guideline for the Management of Heart Failure (December 2015)

Institutional, Voting Ownership/ Organizational, Recusals Committee Speakers Partnership/ Personal or Other Financial Expert By Member Employment Consultant Bureau Principal Research Benefit Witness Section* Clyde W. Northwestern University None None None None None None None Yancy, Chair Feinberg School of Medicine, Division of Cardiology—Professor of Medicine and Chief; Diversity and Inclusion— Vice Dean Mariell Jessup, Fondation Leducq—Chief None None None None None None None Vice Chair Scientific Officer Biykem Baylor College of Medicine, None None None • Novartis None None 7.3.2.10, Bozkurt Department of Medicine— 7.3.2.11, Professor of Medicine; 7.3.3, and Cardiology Section, 9.5. DeBakey VA Medical Center—Chief; The Mary and Gordon Cain Chair & W.A. “Tex” and Deborah Moncrief, Jr.—Chair; Winters Center for Heart Failure Research—Director; Cardiovascular Research Institute—Associate Director Javed Butler Stony Brook University— • Bayer† • Novartis† None • Amgen (DSMB)† None None 7.3.2.10, Division Chief of • Boehringer 7.3.2.11, Cardiology Ingelheim 7.3.3, and 9.5. • CardioCell† Downloaded from http://ahajournals.org by on August 12, 2021 • Luitpold • Medtronic • Merck† • Novartis† • Relypsa† • Takeda • Trevena† • Z Pharma • Zensun Donald E. Thomas Jefferson College None None None None None None None Casey, Jr of Population Health— Faculty; Alvarez & Marsal IPO4Health—Principal and Founder Monica M. University of Michigan— None None None None None None None Colvin Associate Professor of Medicine, Cardiology Mark H. University of Texas None None None None None None None Drazner Southwestern Medical Center—Professor, Internal Medicine Gerasimos S. National and Kapodistrian None None None • Bayer† None None 7.3.2.10, Filippatos University of Athens; • Bayer (DSMB) 7.3.2.11, Attikon University 7.3.3, 9.2, • Novartis† Hospital, Department and 9.5. of Cardiology, Heart • Servier Failure Unit—Professor of Pharmaceuticals† Cardiology • Vifor Gregg C. Ahmanson-UCLA • Amgen None None • Novartis† None None 7.3.2.10, Fonarow Cardiomyopathy Center— • Janssen 7.3.2.11, Director; UCLA Division of Pharmaceuticals 7.3.3, and Cardiology—Co-Chief 9.5. • Novartis† (Continued )

e156 August 8, 2017 Circulation. 2017;136:e137–e161. DOI: 10.1161/CIR.0000000000000509 2017 ACC/AHA/HFSA Heart Failure Focused Update CLINICAL STATEMENTS AND GUIDELINES Appendix 1. Continued

Institutional, Voting Ownership/ Organizational, Recusals Committee Speakers Partnership/ Personal or Other Financial Expert By Member Employment Consultant Bureau Principal Research Benefit Witness Section* Michael M. Brigham and Women's • Merck None None None None None 7.3.2.10, Givertz Hospital—Professor of • Novartis 7.3.2.11, Medicine 7.3.3, and 9.5. Steven M. Cooper University None None None None None None None Hollenberg Hospital—Director, Coronary Care Unit, Professor of Medicine JoAnn Vanderbilt Heart and • Abbott None None • AstraZeneca None None 6.3, Lindenfeld Vascular Institute— • Janssen • Novartis† 7.3.2.10, Director, Advanced Heart Pharmaceuticals 7.3.2.11, Failure and Transplant 7.3.3, 9.5, • Novartis Section—Professor of and 9.6. Medicine • Relypsa† • ResMed† Frederick A. University of Colorado, None None None None None None None Masoudi Anschutz Medical Campus—Professor of Medicine, Division of Cardiology Patrick E. University of Wisconsin None None None None None None None McBride School of Medicine and Public Health—Professor of Medicine and Family Medicine; Associate Director, Preventive Cardiology Pamela N. University of Colorado, None None None None None None None Downloaded from http://ahajournals.org by on August 12, 2021 Peterson Denver Health Medical Center—Associate Professor of Medicine, Division of Cardiology Lynne Warner Brigham and Women’s None None None • Novartis— None None 7.3.2.10, Stevenson Hospital Cardiovascular PARENT trial (PI) 7.3.2.11, Division—Director, • NHLBI— 7.3.3, and Cardiomyopathy and INTERMACS 9.5. Heart Failure Program (Co–PI) Cheryl Azusa Pacific University, None None None None None None None Westlake School of Nursing, Doctoral Programs— Professor

This table represents the relationships of committee members with industry and other entities that were determined to be relevant to this document. These relationships were reviewed and updated in conjunction with all meetings and/or conference calls of the writing committee during the document development process. The table does not necessarily reflect relationships with industry at the time of publication. A person is deemed to have a significant interest in a business if the interest represents ownership of ≥5% of the voting stock or share of the business entity, or ownership of ≥$5000 of the fair market value of the business entity, or if funds received by the person from the business entity exceed 5% of the person’s gross income for the previous year. Relationships that exist with no financial benefit are also included for the purpose of transparency. Relationships in this table are modest unless otherwise noted. According to the ACC/AHA, a person has a relevant relationship IF: a) The relationship or interest relates to the same or similar subject matter, intellectual property or asset, topic, or issue addressed in the document; or b) the company/entity (with whom the relationship exists) makes a drug, drug class, or device addressed in the document, or makes a competing drug or device addressed in the document; or c) the person or a member of the person’s household has a reasonable potential for financial, professional, or other personal gain or loss as a result of the issues/content addressed in the document. *Writing committee members are required to recuse themselves from voting on sections to which their specific relationships with industry and other entities may apply. †Significant relationship. ACC indicates American College of Cardiology; ACCF, American College of Cardiology Foundation; AHA, American Heart Association; DCRI, Duke Clinical Research Institute; DSMB, data safety monitoring board; HFSA, Heart Failure Society of America; NHLBI, National Heart, Lung, and Blood Institute; INTERMACS, Interagency Registry for Mechanically Assisted Circulatory Support; PARENT, Pulmonary artery pressure reduction with entresto; UCLA, University of California, Los Angeles; and VA, Veterans Affairs.

Circulation. 2017;136:e137–e161. DOI: 10.1161/CIR.0000000000000509 August 8, 2017 e157 Yancy et al

Appendix 2. Reviewer Relationships With Industry and Other Entities (Comprehensive)—2017 ACC/AHA/ HFSA Focused Update of the 2013 ACCF/AHA Guideline for the Management of Heart Failure (October 2016)

Institutional, Ownership/ Organizational, Speakers Partnership/ Personal or Other Financial Expert Reviewer Representation Employment Consultant Bureau Principal Research Benefit Witness Kim K. Official University of Houston • Jones & Bartlett None None None None None Birtcher Reviewer— College of Pharmacy— Learning ACC/AHA Clinical Professor Task Force on Clinical Practice Guidelines Akshay S. Official Brigham and Women’s • Medscape None None None • Novartis* None Desai Reviewer— Hospital—Director, Cardiology* • Thoratec HFSA Heart Failure Disease • Merck Management, • Novartis* Advanced Heart Disease Section, • Relypsa* Cardiovascular Division; • St. Jude Associate Professor Medical* of Medicine, Harvard Medical School Anita Official Michael E. DeBakey VA None None None • NIH* • AHA None Deswal Reviewer—AHA Medical Center—Chief • AHA (GWTG of Cardiology; Director, Steering Heart Failure Program; Committee)† Baylor College of • HFSA† Medicine—Professor of Medicine Dipti Official Newport Coast None None None None • St. Jude Medical None Itchhaporia Reviewer— Cardiology—Robert ACC Board of and Georgia Roth Trustees Endowed Chair for Excellence in Cardiac Downloaded from http://ahajournals.org by on August 12, 2021 Care; Director of Disease Management Ileana L. Piña Official Montefiore Medical • Relypsa None None None None None Reviewer—AHA Center—Associate Chief for Academic Affairs, Cardiology; Professor of Medicine & Epidemiology and Population Health— Albert Einstein College of Medicine

Geetha Official University of Missouri- None None None None None None Raghuveer Reviewer— Kansas City School of ACC Board of Medicine—Professor of Governors Pediatrics; Children's Mercy Hospital— Pediatric Cardiology James E. Official Tufts Medical • Lantheus None None • Gilead (DSMB) • Abbott None Udelson Reviewer— Center—Chief, Division Medical Imaging • GlaxoSmithKline Laboratories HFSA of Cardiology (DSMB) • AHA* • NHLBI • Circulation/ • Otsuka Circulation: Heart Failure† • HFSA (Executive Council)† • Pfizer/ GlaxoSmithKline • Sunshine Heart Mary Norine Official St Vincent Heart None None None None • Corvia Medical None Walsh Reviewer— Center of Indiana— • Otsuka ACC Board of Medical Director, Heart • PCORI Trustees Failure and Cardiac Transplantation • Thoratec (Continued )

e158 August 8, 2017 Circulation. 2017;136:e137–e161. DOI: 10.1161/CIR.0000000000000509 2017 ACC/AHA/HFSA Heart Failure Focused Update CLINICAL STATEMENTS AND GUIDELINES Appendix 2. Continued

Institutional, Ownership/ Organizational, Speakers Partnership/ Personal or Other Financial Expert Reviewer Representation Employment Consultant Bureau Principal Research Benefit Witness David A. Organizational Newark Beth Israel • Maquet • Novartis None • XDx* None None Baran Reviewer— Medical Center— • Otsuka* • NIH* ISHLT Director of Heart Failure and Transplant Research Kenneth Organizational Wm. S. Middleton None None None None • CHEST None Casey Reviewer— Memorial Veterans CHEST Hospital—Director, Sleep Medicine M. Fuad Jan Organizational Aurora Advanced None None None None None None Reviewer— Healthcare— CHEST Cardiologist Kenneth W. Organizational Georgetown University None None None None None None Lin Reviewer— School of Medicine— AAFP Clinician Educator Track, Associate Professor Joaquin E. Content Oregon Health & None None None None • ACC/AHA† None Cigarroa Reviewer— Science University— • AHA† ACC/AHA Clinical Professor of • ASA† Task Force on Medicine Clinical Practice • Catheterization Guidelines and Cardiovascular Intervention† • NIH • Portland Metro Area AHA (President)† • SCAI Quality Interventional

Downloaded from http://ahajournals.org by on August 12, 2021 Council† Lee A. Content University of • Blue Cross/Blue None None • Johns Hopkins • Association None Fleisher Reviewer— Pennsylvania Health Shield* (DSMB) of University ACC/AHA System—Robert • NQF† Anesthesiologists† Task Force on Dunning Dripps Professor • Yale University • NIH Clinical Practice of Anesthesiology Guidelines and Critical Care; Chair, Department of Anesthesiology & Critical Care Samuel S. Content Nemours/Alfred I. • FH Foundation† None None • FH None None Gidding Reviewer— duPont Hospital for • International FH Foundation† ACC/AHA Children—Chief, Foundation† • NIH* Task Force on Division of Pediatric Clinical Practice Cardiology Guidelines James L. Content Massachusetts General • Critical None None • Amgen (DSMB) None None Januzzi Reviewer Hospital—Hutter Diagnostics* • Boeringer Family Professor of • Novartis* Ingelheim Medicine in the Field of • Phillips (DSMB)* Cardiology • Roche • Janssen Diagnostics* Pharmaceuticals (DSMB) • Sphingotec* • Prevencio* José A. Content UT Southwestern None None None None None None Joglar Reviewer— Medical Center— ACC/AHA Professor of Task Force on Internal Medicine; Clinical Practice Clinical Cardiac Guidelines Electrophysiology— Program Director Edward K. Content Johns Hopkins None None None None None None Kasper Reviewer Cardiology—E. Cowles Andrus Professor in Cardiology (Continued ) Circulation. 2017;136:e137–e161. DOI: 10.1161/CIR.0000000000000509 August 8, 2017 e159 Yancy et al

Appendix 2. Continued

Institutional, Ownership/ Organizational, Speakers Partnership/ Personal or Other Financial Expert Reviewer Representation Employment Consultant Bureau Principal Research Benefit Witness Wayne C. Content University of • Abbott None None • NIH • Amgen* None Levy Reviewer Washington— Laboratories • Novartis* • AHA Professor of Medicine • Biotronik • St. Jude • HeartWare* • GE Healthcare Medical* • Novartis* • HeartWare • Resmed* • PharminIN • Thoratec Judith E. Content SUNY Downstate None None None None • Association None Mitchell Reviewer Medical Center— of Black Director/Heart Cardiologists† Failure Center; SUNY Downstate College of Medicine—Associate Professor of Medicine Sean P. Content Mount Sinai School of • Acorda None None • Thoratec† None None Pinney Reviewer—ACC Medicine—Associate Therapeutics • NIH† Heart Failure Professor of Medicine, • Thoratec and Transplant Cardiology • XDx Council Randall C. Content Cleveland Clinic • BioControl None None • Medtronic • St. Jude Medical None Starling Reviewer—ACC Department of • Medtronic • NIH* Heart Failure Cardiovascular • Novartis • Novartis† and Transplant Medicine—Vice Council Chairman, Department of • St. Jude Cardiovascular Medicine; Medical† Section Head,Heart Failure & Cardiac Transplant W.H. Wilson Content Cleveland Clinic None None None • NIH* • Alnylam None

Downloaded from http://ahajournals.org by on August 12, 2021 Tang Reviewer Foundation—Assistant Pharmaceuticals Professor of Medicine • NIH • NHLBI • Roche • Novartis • Thoratec Emily J. Tsai Content Columbia University None None None • Bayer† None None Reviewer College of Physicians • Bristol-Myers & Surgeons—Assistant Squib† Professor of Medicine, • NHLBI* Division of Cardiology Duminda N. Content Li Ka Shing Knowledge None None None • CIHR (DSMB)† None None Wijeysundera Reviewer— Institute of St. • CIHR* ACC/AHA Michael’s Hospital— • Heart and Stroke Task Force on Scientist; University Foundation of Clinical Practice of Toronto—Assistant Canada* Guidelines Professor, Department of Anesthesia and • Ministry of Institute of Health Health & Long- Policy Management term Care of and Evaluation Ontario* • PCORI DSMB)†

This table represents the relationships of reviewers with industry and other entities that were disclosed at the time of peer review, including those not deemed to be relevant to this document. The table does not necessarily reflect relationships with industry at the time of publication. A person is deemed to have a significant interest in a business if the interest represents ownership of ≥5% of the voting stock or share of the business entity, or ownership of ≥$5000 of the fair market value of the business entity, or if funds received by the person from the business entity exceed 5% of the person’s gross income for the previous year. A relationship is considered to be modest if it is less than significant under the preceding definition. Relationships that exist with no financial benefit are also included for the purpose of transparency. Relationships in this table are modest unless otherwise noted. Names are listed in alphabetical order within each category of review. American College of Physicians did not provide a peer reviewer for this document. *Significant relationship. †No financial benefit. AAFP indicates American Academy of Family Physicians; ACC, American College of Cardiology; ACCF, American College of Cardiology Foundation; AHA, American Heart Association; ASA, American Stroke Association; CHEST, American College of Chest Physicians; CIHR, Canadian Institutes of Health Research; DSMB, data safety monitoring board; FH, familial hypercholesterolemia; GWTG, Get With The Guidelines; HFSA, Heart Failure Society of America; ISHLT, International Society for Heart and Lung Transplantation; NIH, National Institutes of Health; NHLBI, National Heart, Lung, and Blood Institute; NQF, National Quality Forum; PCORI, Patient-Centered Outcomes Research Institute; SCAI, Society for Cardiac Angiography and Interventions; SUNY, State University of New York; UT, University of Texas; and VA, Veterans Affairs.

e160 August 8, 2017 Circulation. 2017;136:e137–e161. DOI: 10.1161/CIR.0000000000000509 2017 ACC/AHA/HFSA Heart Failure Focused Update CLINICAL STATEMENTS AND GUIDELINES Appendix 3. Abbreviations

ACE = angiotensin-converting enzyme ARB = angiotensin-receptor blocker ARNI = angiotensin receptor–neprilysin inhibitor BNP = B-type natriuretic peptide

BP = blood pressure COR = Class of Recommendation CPAP = continuous positive airway pressure EF = ejection fraction GDMT = guideline-directed management and therapy HFpEF = heart failure with preserved ejection fraction HFrEF = heart failure with reduced ejection fraction LOE = Level of Evidence LVEF = left ventricular ejection fraction NT-proBNP = N-terminal pro-B-type natriuretic peptide QoL = quality of life RCT = randomized controlled trial Downloaded from http://ahajournals.org by on August 12, 2021

Circulation. 2017;136:e137–e161. DOI: 10.1161/CIR.0000000000000509 August 8, 2017 e161 August 16, 2021

Clarance S. Farmer, II, M.D.

Urologic

Presentation: Men’s Health Update: Pros and Cons of Testosterone Replacement; What is the Value of a Screening PSA?

References:

1. https://www.auanet.org/#

2. https://seer.cancer.gov/statfacts/html/prost.html

3. https://www.uptodate.com/contents/screening-for-prostate-cancer#H2191332858

4. https://cancerstatisticscenter.cancer.org/#!/cancer-site/Prostate

5. https://www.cancer.org/cancer/prostate-cancer-about/key-statistics.html

6. https://gis.cdc.gov/Cancer/USCS/#/AtAGlance/

7. https://uspreventiveservicestaskforce.org/uspstf/recommendation/prostate-cancer-screening

The Role of Botulinum ToxininVasospastic Disorders of the Hand

Michael W. Neumeister, MD, FRCSC, FACS

KEYWORDS  Vasospasm  Raynaud  Botox  Botulinum toxin  Vasodilation  Pain  Ischemia

KEY POINTS  At present, botulinum injections offer a low-risk method for the treatment of symptomatic Raynaud phenomenon.  Although the mechanism has not been fully elucidated, this finding provides patients with another means of medical management before moving to sympathectomy.  Repeated injections may be required.

INTRODUCTION of botulinum toxin for the treatment of patients with Raynaud phenomenon. The hand is a true end organ with an arborization of vasculature that continues from the palms to VASCULAR INSUFFICIENCY FROM the fingertips and nail beds. A multitude of intrinsic VASOSPASM OF THE DIGITAL VESSELS and extrinsic factors can affect the normal anat- omy and physiology of the blood vessels to the Vasospastic disorders of the hand are a common hand. Blood vessels may be affected by trauma, phenomenon and are manifest by remitting and re- inflammation, infection, autoimmune disorders, lapsing clinical symptoms. Recurring episodes of pharmacologic agents, neoplasms, and endocrine vasospasm become problematic as the frequency abnormalities, as well as other factors that may and duration of the episodes increase. One of the affect the central nervous system, the peripheral earliest descriptions of vasospasm of digital ar- nervous system, and the neuropeptides that con- teries was by Maurice Raynaud in 1862, who trol vascular tone. Vasospasm, which is defined described the process as “A local asphyxia of as inappropriate tone of the arteries or veins in the extremities as the result of increased irritability the hand, can result in impaired vasodilatation, of the central parts of the cord presiding over cold intolerance, and digital ischemia. It is esti- vascular innervation.”5,6 The vasospastic disorder mated that more than 9 million people have vaso- characterized by Maurice Raynaud was called spasm of the digits annually.1–4 Women are more Raynaud disease for decades until John Hutchin- susceptible than men at a ratio of 9:1 and young son in 1901 correctly articulated that there were adults experience more vasospastic episodes many causes for vasospasm of the digital vessels that those more than 40 years of age.1–4 Although and that the term “disease” was probably not advances have been made in the last decade, un- appropriate for this condition. Sir Thomas Louis derstanding of digital vasospasm, also known as was the first to describe the physiologic features Raynaud phenomenon, is still limited. This article to distinguish primary Raynaud phenomenon focuses on vasospasm of the hand and the use from secondary Raynaud conditions.5 Primary

The author is a consultant for MTF. Institute for Plastic Surgery, Southern Illinois University School of Medicine, 747 North Rutledge, 3rd Floor, PO Box 19653, Springfield, IL 62794, USA E-mail address: [email protected]

Hand Clin 31 (2015) 23–37 http://dx.doi.org/10.1016/j.hcl.2014.09.003

0749-0712/15/$ – see front matter Ó 2015 Elsevier Inc. All rights reserved. hand.theclinics.com Downloaded for Anonymous User (n/a) at The University of Tennessee Health Science Center from ClinicalKey.com by Elsevier on August 17, 2021. For personal use only. No other uses without permission. Copyright ©2021. Elsevier Inc. All rights reserved. 24 Neumeister

Box 1 Many medical conditions or illnesses may have one of Conditions associated with secondary Raynaud their manifestations present as Raynaud. The many potential causes or related entities that are associated Rheumatologic disease with secondary Raynaud phenomenon are indicated here. Systemic sclerosis Adapted from Bakst R, Merola JF, Franks AG Jr, et al. Raynaud’s phenomenon: pathogenesis and manage- Systemic lupus erythematosus ment. J Am Acad Dermatol 2008;59(4):635; with Mixed connective tissue disease permission. Dermatomyositis Rheumatoid arthritis Sjo¨ gren syndrome Raynaud phenomenon does not have any other Vasculitis associated medical condition linked to the vaso- Vascular occlusive disease spastic disorder. In secondary Raynaud phenom- enon, the vasospastic disorder is linked to other Buerger disease systemic disorders (Box 1, Table 1).1,6 Arteriosclerosis The term Raynaud phenomenon is used here to Thromboembolic disease describe the clinical picture of pallor, cyanosis, and rubor. The disorder is the result of vascular Drug induced embarrassment to the hand from dysfunction of Amphetamines the homeostatic neural control of blood vessels. Beta-adrenergic blockers The resultant paroxysmal vasoconstriction com- bined with episodes of vasodilation manifest in Bleomycin the color changes noted earlier, pain, and ulti- Cisplatin mately ulceration of the fingertips. The histories Cyclosporin obtained from patients often describe their fingers Ergots as being blanched and cool, which is especially brought on by cold exposure or stress (Fig. 1). Interferon alfa There are episodes of secondary vasodilation in Vinblastine which the digit becomes more swollen and fusi- Hematologic syndromes form with characteristic dysesthesias. The longer and more severe the vasospasm, the more likely Polycythemia the patient is to experience fingertip pain and Paraproteinemia possible ulceration. Persistent ischemia may lead Cryoglobulinemia to deep ulceration, bone exposure, and 7,8 Cryofibrinogenemia amputation. The physical examination findings on patients Cold agglutinin disease with Raynaud disease or phenomenon may Homocysteinemia appear normal when a vasospastic episode has Protein C, protein S, antithrombin III not occurred recently. The surgeon should look deficiency for signs of ulceration, inflammation, dry gangrene, Factor V Leiden discoloration, nail bed petechiae, poor turgor, or dystrophic nails or fingertips. The capillary refill Environmental associations may be normal or sluggish but the Allen test is usu- Vibration injury ally normal. Frostbite Tests used to evaluate the vascularity of the hand are included in Box 2. The management of Anatomic syndromes Raynaud phenomenon is variable depending on Scalenus anticus syndrome the cause and the severity of the condition. Essen- Cervical rib tial conservative measures include smoking cessation, maintaining a warm environment, and Infectious causes reducing stress. These early measures are often Hepatitis B and C (associated with enough for those patients who have mild or very cryoglobulinemia) infrequent bouts of vasoconstriction. When vaso- Mycoplasma infections (cold agglutinins) spastic disorders result in prolonged symptomatic problems for patients, drug therapy is indicated. Parvovirus B19 An extensive array of medications has been used

Downloaded for Anonymous User (n/a) at The University of Tennessee Health Science Center from ClinicalKey.com by Elsevier on August 17, 2021. For personal use only. No other uses without permission. Copyright ©2021. Elsevier Inc. All rights reserved. Vasospastic Disorders of the Hand 25

Table 1 Box. 2 Different characteristics of primary versus Tests used to evaluate perfusion in the hand secondary Raynaud phenomenon Doppler Characteristics Primary Secondary Doppler ultrasonography Associated No Yes Digital plethysmography (pulse volume autoimmune recordings) disease Radial-brachial index 5 1.0 Age at <30 >30 5 onset (y) Digital-brachial index 1.0 Pain with Infrequent Frequent Less than 0.7 is abnormal, compromising heal- attacks ing of ulcers, wounds Finger Symmetric Asymmetrical Less than 0.5 indicates impending cell death involvement (hand or digital ischemia) Nail fold Normal Dilated From 0.7 to 1.0 indicates arterial flow capillaries with vessel compromise dropout Color duplex imaging Onycholysis Infrequent Frequent and/or Oxygen saturation pterygium Greater than 90% on room air; compare with inversum contralateral side Tuft pits and/or Infrequent Frequent Skin surface temperature determination necrosis (w30C) Autoantibodies Negative or Increased Laser Doppler fluxmetry low titers titers Cold stress testing Primary Raynaud phenomenon is not associated with any Hands exposed to cool air or water (5–8C) condition. Secondary Raynaud phenomenon is related to other conditions or diseases. Laser Doppler perfusion imaging Adapted from Bakst R, Merola JF, Franks AG Jr, et al. Ray- naud’s phenomenon: pathogenesis and management. J Vital capillaroscopy Am Acad Dermatol 2008;59(4):635; with permission. Scintigraphy Magnetic resonance angiography Computed tomography angiography for vasospastic disorders. Each medication has Many investigations can be used to help define normal and abnormal vascularity and perfusion of been used to block a certain part of the chain the digits. of events thought to be involved with vasospasm of the digital vessels. Box 3 lists the mediators of vascular tone in the hand. The first-line drug used in Raynaud phenomenon is usually a calcium channel blocker to induce vasodilation of the digi- tal vessels.1,6,9–35 Table 2 shows the various phar- macologic treatments to manage Raynaud phenomenon.36–40 Many medications have been tried for this problematic disorder, which indicates that there is no gold standard or so-called magic bullet for the treatment of vasospasm in the hand. Many studies have shown some benefit with the use of these medications, although recal- citrant vasospasm still occurs in many patients, especially those with secondary Raynaud phe- nomenon.1,41 Recent, promising studies have focused on phosphodiesterase 5 (PDE 5) inhibitors as a potential oral treatment of Raynaud.42–50 Lin Fig. 1. Raynaud phenomenon characterized by digital and colleagues51 reviewed the role of PDE 5 in ischemia, pain, and occasional ulcerations. vascular homeostasis. They noted that the human

Downloaded for Anonymous User (n/a) at The University of Tennessee Health Science Center from ClinicalKey.com by Elsevier on August 17, 2021. For personal use only. No other uses without permission. Copyright ©2021. Elsevier Inc. All rights reserved. 26 Neumeister

Box 3 PDE 5 degrades cyclic GMP and therefore limits Mediators of the vascular tone of the hand vasodilation within the physiologic norm. PDE 5 inhibitors prevent the degradation of cyclic GMP  Sympathetic nervous system to promote long-lasting vasodilatation of vessels.  Alpha adrenergic receptors of smooth mus- Although most commonly used for erectile dys- cle (alpha 2c) function, PDE 5 inhibitors are being used for other vascular disorders including Raynaud phenome- - Norepinephrine non. Numerous case reports (level V evidence)  Parasympathetic nervous system and case series (level IV evidence) have shown im-  Central nervous system provements with respect to the number and severity of the spastic attacks in patients treated  Nitric oxide/cyclic GMP system with PDE 5 inhibitors, but 2 of 3 randomized,  Reactive oxygen species controlled, double-blinded studies (level II evi-  Redox pathways dence) have failed to show any significant 42,50,52–54  Rho kinase benefit.  Estrogen BOTULINUM TOXIN  Substance-P  Calcitonin gene–related peptide In 1897, Emile van Ermengen, a microbiologist at  Neuropeptide Y the University of Ghent, Belgium, discovered that the bacterium Clostridium botulinum was respon-  Vasoactive intestinal peptide sible for producing the protein botulinum toxin.5  Glutamate C botulinum is a gram-negative, spore-forming,  Prostacyclin (prostaglandin I2) obligate anaerobic bacterium and is considered one of nature’s most potent neuroparalytic agents.  Endothelium-derived hyperpolarizing factor P. Tessmer Snipe and Hermann Sommer first  Endothelium 1 purified the toxin in 1928.55 The first elucidation  Thromboxane of the mechanism of action of botulinum toxin as  Angiotensin an inhibitor of acetylcholine release at neuromus- cular junctions was described in 1949 by Arnold  Platelet activation Burgen.56 There are 8 serologically different types The vessels of the hand and digits respond to stimuli of botulinum toxin (A–H). The protein consists of a from various sources, including nerve endings, humor- light chain (50 kDa) and a heavy chain (100 kDa) al factors, tissue activation, and intrinsic factors. The joined by a disulfide bond.57 Botulinum toxin is multifactorial influence signals the balance of vaso- construction and vasodilation. considered one of the most lethal toxins known, with a median lethal dose of 1.3 to 2.1 ng/kg when administered intravenously or intramuscu- larly. One gram of botulinum toxin has the potential to kill 1 million people. However, in limited doses, genome encodes 21 phosphodiesterase genes. botulinum toxin A and B have been used for ther- Furthermore the genes could be grouped into 11 apeutic reasons. The US Food and Drug Adminis- families with almost 100 isoforms. PDE 5 and its tration (FDA) has approved 8 different uses for 3 isoforms A1, A2, and A3 are the most studied botulinum toxin A (Box 4), but numerous off-label phosphodiesterase enzymes that play a role in pe- uses have been documented in the literature as ripheral vascular control. well (Box 5).57,58 PDE 5 is affects vasodilatation through its effects Botulinum toxin’s most commonly known mech- on guanosine monophosphate (GMP) and the nitric anism of action is at the nerve ends of the neuro- oxide cycle. Nitric oxide, a natural vasodilator, has muscular junction.59 The toxin is taken up by 3 isoforms of nitric oxide synthase: inducible nitric the nerve endings. The light chain of the toxin is oxide synthase, endothelial nitric oxide synthase, an enzyme that cleaves a membrane-bound pro- and neural nitric oxide synthase. Nitric oxide acti- tein, called synaptosomal-associated protein 25 vates guanylate cyclase to increase cyclic GMP (SNAP-25), on the acetylcholine-filled vesicles. levels, which induce smooth muscle relaxation ul- This cleavage prevents transport and exocytosis timately producing vasodilation of blood vessels. of the acetylcholine. The inhibition of acetylcholine Pathologic imbalance of nitric oxide levels may in- release across the neuromuscular junction pre- fluence the normal physiologic regulation of vents muscle contraction, clinically manifesting vascular tone in the digits. as paralysis. However, the clinical manifestations

Downloaded for Anonymous User (n/a) at The University of Tennessee Health Science Center from ClinicalKey.com by Elsevier on August 17, 2021. For personal use only. No other uses without permission. Copyright ©2021. Elsevier Inc. All rights reserved. Vasospastic Disorders of the Hand 27

Table 2 Pharmacologic treatment of Raynaud phenomenon

Medication Dosage Comments Calcium Channel Blockers Nifedipine 10–30 mg TID Least cardioselective, most efficacious class Nifedipine, sustained release 30–90 mg QD Sustained release preparation reduces side effects Felodipine 2.5–10 mg QD — Amlodipine 2.5–10 mg QD — Diltiazem, sustained release 120–33 mg QD Intermediate cardioselectivity, minimal controlled data Verapamil 40–120 mg TID Most cardioselective, benefit not shown Vasodilators Nitroglycerin 0.25–0.5 in topical 1% Benefit shown; significant side effects avoided Nitroglycerin 0.25–0.5 in topical 2% Significant side effects from systemic absorption L-Arginine 2–8 g QD Useful in patients with low blood pressure Sodium nitroprusside 0.3 mg/kg/min/IV Used in critical ischemia Prostaglandins Intravenous iloprost 0.5–2.0 ng/kg/min IV for 1–5 d Used in critical ischemia Oral iloprost 50–100 mg BID Conflicting data regarding oral prostaglandins Beraprost 60 mg TID Conflicting data regarding oral prostaglandins Epoprostenol 0.5–6 ng/kg/min for 2–5 d Used in critical ischemia Phosphodiesterase Inhibitors Sildenafil 50 mg BID One placebo-controlled trial indicated benefit Cilostazol 50–100 mg BID One placebo-controlled trial indicated benefit ACE Inhibitors Enalapril 2.5–20 mg QD Few controlled data Captopril 6.5–25 mg TID Few controlled data Losartan 50 mg QD One placebo-controlled trial indicated benefit Selective Serotonin Reuptake Inhibitors Fluoxetine 20 mg QD One pilot study indicated benefit Endothelin Receptor Antagonists Bosentan 62.5–125 mg BID Reduces number of new ulcers but no benefit shown compared with existing ones; hepatotoxic (continued on next page)

Downloaded for Anonymous User (n/a) at The University of Tennessee Health Science Center from ClinicalKey.com by Elsevier on August 17, 2021. For personal use only. No other uses without permission. Copyright ©2021. Elsevier Inc. All rights reserved. 28 Neumeister

Table 2 (continued)

Medication Dosage Comments Sympatholytics Prazosin 1–5 mg BID Efficacy transient, may wane over time Guanethidine 10–50 mg QD Few controlled data Phenoxybenzamine 10 mg QID Few controlled data Rho Kinase Inhibition Fasudil 30 mg/min for 30 min IV Benefit shown in treatment of severe pulmonary hypertension Anticoagulation/Antithrombotic Low-molecular-weight heparin — Used in critical ischemia Warfarin — Few controlled data Aspirin 81–325 mg QD Few controlled data Dipyridamole 75–100 mg QID Few controlled data Tissue plasminogen activator — Used in critical ischemia Antioxidants Probucol 500 mg QD Few controlled data Botulinum toxin 40–100 U/hand Improve perfusion, decrease pain Miscellaneous Vitamin E 500 U QD Useful in patients with low blood pressure Slo-Niacin 250 mg QD Useful in patients with low blood pressure Griseofulvin (microsize) 500 mg daily Useful in patients with low blood pressure Pentoxifylline 400 mg TID Useful in patients with low blood pressure

Many drugs have been used to treat Raynaud to inhibit vasospasm and improve blood flow. Abbreviations: ACE, angiotensin-converting enzyme; BID, twice a day; IV, intravenous; QD, daily; QID, 4 times a day; TID, 3 times a day. Adapted from Bakst R, Merola JF, Franks AG Jr, et al. Raynaud’s phenomenon: pathogenesis and management. J Am Acad Dermatol 2008;59(4):643; with permission.

have a delay of onset of 1 to 4 days as the toxin is Box 4 FDA-approved botulinum toxin products taken up and is able to cleave the SNAP-25 pro- tein. The duration of action is 2 to 4 months, allow- Indication ing time for the SNAP-25 proteins to regenerate. Cervical dystonia However, botulinum toxin has other mechanisms of action. Axillary hyperhidrosis Strabismus THE ROLE OF BOTULINUM TOXIN IN Blepharospasm VASOSPASTIC DISORDERS OF THE HAND Urinary incontinence Botulinum toxin has been used to treat vasospasm Migraine headache of digital vessels for more than a decade. The toxin Upper limb spasticity is injected directly into the hand in patients who Torticollis have primary or secondary Raynaud phenomenon (Fig. 2). Botulinum toxin is indicated for patients

Downloaded for Anonymous User (n/a) at The University of Tennessee Health Science Center from ClinicalKey.com by Elsevier on August 17, 2021. For personal use only. No other uses without permission. Copyright ©2021. Elsevier Inc. All rights reserved. Vasospastic Disorders of the Hand 29

Box 5 Off-label uses for botulinum toxin products

Achalasia Back pain Carpal tunnel syndrome Cerebral palsy–related limb spasticity Chronic anal fissure Chronic pain Crow’s feet Delayed gastric emptying Fig. 2. The reconstituted onabotulinumtoxinA is in- Dystonia jected around each neurovascular bundle. Approxi- Enlarged prostate mately 10 units of onabotulinumtoxinA (2 mL) are used to bathe the neurovascular bundle. (From Neu- Epilepsy meister MW. Botulinum toxin type a in the treatment Epiphora of Raynaud’s phenomenon. Journ Hand Surg 2010;35(12):2087; with permission.) Essential tremor Fibromyalgia Hereditary paraplegia Interstitial cystitis injection of botulinum toxin, perfusion improves Ischemic digits to the digits and the pain associated with the Ray- naud resolves almost immediately for most pa- Lateral epicondylitis tients, even those patients who have Piriformis syndrome ulcerations.38–40,60–62 Pruritus Observations that injections of botulinum toxin Raynaud phenomenon may diminish or even ameliorate chronic pain and improve perfusion in patients with Raynaud Rhinitis have sparked significant basic science research Sialorrhea but still the mechanism of action of botulinum toxin Spasmodic dysphonia is not yet fully elucidated. It seems that botulinum toxin may inhibit the activities of sympathetic and Stuttering sensory nerves, neurotransmitters, and signal Tardive dyskinesia transduction pathways.57,58,63–65 It is likely that Tourette syndrome botulinum toxin acts at several sites on vessels Vaginismus and nerves. Vasospasm not only diminishes blood flow but also causes pain. Recurring episodes of Adapted from Neumeister, Webb KN, Romanelli M. the Raynaud phenomenon lead to chronic pain. Minimally invasive treatment of Raynaud’s phenome- Chronic pain is propagated by the C-fibers. Injured non: the role of botulinum type A. Hand Clinics 2014;30:17–24; with permission. C-fiber nociceptors become sensitized and un- dergo significant chemical and biochemical alter- ations. Adrenergic receptors are upregulated in the sensitized C-fibers. Sympathetic stimulation who have failed conventional treatment (see Ta- results in a release of norepinephrine, which acts ble 2). Each drug used for Raynaud vasospasm on the adrenergic receptors to produce both vaso- targets different physiologic sites to induce vaso- constriction and pain.66 Stress and cold each dilation or improve blood flow to the digits. The re- respond to sympathetic discharges and are vaso- sults of this pharmacologic manipulation vary spasm stimuli that result in pain in patients with widely. In the past, patients who did not respond Raynaud.64,66,67 The pain may also be the result to these medications underwent surgical interven- of stimulation of the release of neurotransmitters tion such as sympathectomies. However, surgery such as substance P, which acts as a central ner- has the potential for complications and has not vous system neurotransmitter. Pain nociceptors been uniformly successful; botulinum toxin can release substance P and glutamate. Botulinum now offer these patients an additional option. Mul- has been shown to block both substance P and tiple case series have shown that, following glutamate.

Downloaded for Anonymous User (n/a) at The University of Tennessee Health Science Center from ClinicalKey.com by Elsevier on August 17, 2021. For personal use only. No other uses without permission. Copyright ©2021. Elsevier Inc. All rights reserved. 30 Neumeister

Fonseca and colleagues6 recently published THE CLINICAL USE OF BOTULINUM TOXIN their hypothesis for the mechanism of action of FOR VASOSPASM botulinum toxin, in particular its ability to block vasoconstriction related to cold sensitivity. Botulinum toxin holds significant promise for (Fig. 3) The toxin interacts with Rho/Rho kinase treating vasospasm. The various case series of patients treated with botulinum toxin are listed in to inactivate the Rho kinase. This interaction leads 38–40,59,61 to inhibition of smooth muscle contraction through Table 3. From 75% to 100% of patients interference with Ca21 sensitivity and nitric oxide. report significant improvement of the pain associ- Laboratory studies on flap viability, vasodilation, ated with the symptoms of Raynaud phenomenon. and tissue perfusion have shown improvement in Up to 37% of patients have required a repeat injec- tion. Some patients have ongoing relief years after tissue perfusion in ischemic models and vessel 40 diameter. In addition, numerous vascular studies the initial injection. Fregene and colleagues found in animal models have shown the value of botuli- that 75% of patients experienced a significant num toxin in inducing vasodilation or preventing reduction in pain, 56% experienced improvement thrombosis.68–73 Improved tissue or flap viability in transcutaneous oxygen saturation, 48% of ul- has similarly been documented with the use botu- cers healed within 9.5 weeks, and 89% of patients linum toxin.74–78 Improved tissue viability is pre- experienced improvements after 1 treatment. sumably through vasodilation and augmented The indications and contraindications for the cellular perfusion in these studies. use of botulinum toxin for vasospastic disorders

Fig. 3. Pathways involved in vasoconstrictions induced by cold or stress in Raynaud phenomenon. Nitric oxide is a natural vasodilator, and also increases intracellular Ca21. However, as the skin is exposed to cold in patients with Raynaud phenomenon, the vascular smooth muscle generates reactive oxygen species that are involved in the activation of the RhoA kinase pathway to increase the activity of actin and myosin, leading to vasoconstriction. Furthermore, the activation of the RhoA kinase pathway induces the translocation of a2c adrenergic receptor to the cell surface, where they are stimulated by norepinephrine to create vasoconstriction. Botulinum toxin may block the production of reactive oxygen species, preventing the activation of the RhoA kinase pathway and permitting the unopposed action of nitric oxide leading to vasodilation. Botulinum toxin may alternatively act to increase intracellular Ca21 leading to vasodilation. This pathway is theoretic, based on previous work docu- mented with the RhoA kinase pathway (red arrows are inhibitory; thin black arrows are stimulatory). (Adapted from Fonseca, Abraham D, Ponticos M. Neuronal regulators and vascular dysfunction in Raynaud’s phenomenon and systemic sclerosis. Current Vasc Pharm 2009,7(1):36.)

Downloaded for Anonymous User (n/a) at The University of Tennessee Health Science Center from ClinicalKey.com by Elsevier on August 17, 2021. For personal use only. No other uses without permission. Copyright ©2021. Elsevier Inc. All rights reserved. Downloaded for Anonymous User(n/a) at The University ofTennessee Health ScienceCenter from ClinicalKey.comby Elsevier on August 17,2021. For personal use only. No other uses without permission. Copyright ©2021. Elsevier Inc. All rights reserved.

Table 3 Summary of findings from retrospective studies using injected botulinum toxin A to treat patients with Raynaud with ischemic digits

Average Length Average Duration Number of Number of Average Percentage with of Follow-up in of Pain Relief in Patients (Male/ Patients with Age in Years Symptomatic Months Months Number of Complications Study Female) 1/2 Raynaud (Range) Relief (Number) (Maximum) (Range) (%) Sycha et al,62 2 (0/2) 1/1 50.5 (19–63) 100 (2/2) 1.75 (2) 1.75 (1.5–2) None 2004 Van Beeket al,39 11 (2/9) 1/10 50.8 (23–70) 100 (11/11) 9.6 (30) Not reported Temporary intrinsic muscle 2007 weakness, 3 (27) Fregene 26 (12/14) 15/11 55 (37–72) 75 (19/6) 18 (45) Not Reported Temporary intrinsic muscle et al,40 2009 weakness, 6 (23)

Transient dysesthesia, Hand the of Disorders Vasospastic injected distal, 1 (4) Neumeister 19 (7/12) 13/6 44.1 (15–72) 84 (16/19) Not reported 23.4 (0.5–59) Temporary intrinsic muscle et al,38 2009 (59) weakness, 3 (16) Neumeister,60 33 (14/19) 23/10 43.0 (15–72) 84 (28/33) (103) (0.5–103) Temporary intrinsic muscle 2010 weakness, 7 (21%) Adapted from Neumeister, Webb KN, Romanelli M. Minimally invasive treatment of Raynaud’s phenomenon: the role of botulinum type A. Hand Clinics 2014;30:22; with permission. 31 32 Neumeister

Box 6 Box 7 Indications for botulinum toxin in patients with The contraindications for the use of botulinum ischemic digits toxin A

 Raynaud phenomenon  Known allergies  Vascular insufficiency not amenable to bypass  Pregnancy surgery  Breast-feeding mothers Adapted from Neumeister MW, Webb KN, Romanelli  Active infections M. Minimally invasive treatment of Raynaud phenom- enon – the role of botulinum type A. Hand Clinics  Myasthenia gravis 2014;30;1:17–24; with permission.  Patients on medications that decrease neuro- muscular transmission:  Calcium channel blockers  Penicillamine  are listed in Boxes 6 and 7 respectively. Clinical Aminoglycosides examples are included to show the efficiency of  Lincosamides botulinum (Figs. 4–8) There are still several ques-  Polymyxins tions to be answered with the use of botulinum  Magnesium sulfate toxin in vasospastic disorders. The exact dose response curve has not been elucidated, meaning  Anticholinesterases that lower doses might suffice. In contrast,  Succinylcholine perhaps greater doses are required for those pa-  Quinidine tients who did not respond as well as expected. Understanding the mechanism of action of botuli- Adapted from Neumeister MW, Webb KN, Romanelli num toxin is important because it may help clini- M. Minimally invasive treatment of Raynaud phenom- enon – the role of botulinum type A. Hand Clinics cians to define which patients are going to 2014;30;1:17–24; with permission. respond better than others to the toxin injections.

Fig. 4. A patient with recalcitrant Raynaud and an ischemic ulcer on the index finger (A). A total of 100 units of onabotulinumtoxinA were injected into the hand. Immediate relief of pain was reported and the perfusion improved. The ulcers healed except for an area of exposed distal phalanx that required a revision amputation (B–D). The patient remains symptom free at 6-year follow-up.

Downloaded for Anonymous User (n/a) at The University of Tennessee Health Science Center from ClinicalKey.com by Elsevier on August 17, 2021. For personal use only. No other uses without permission. Copyright ©2021. Elsevier Inc. All rights reserved. Vasospastic Disorders of the Hand 33

Fig. 5. Patient with Raynaud who underwent onabotulinumtoxinA injections (A). There is observable improve- ment in perfusion after the injections (B).

Fig. 6. Laser scans before perfusion of a patient with Raynaud injected with 100 units of botulinum toxin type A (A). A marked increase in perfusion was observed after the injection (B).

Fig. 7. Injection of 50 units of onabotulinumtoxinA per hand. Preinjection (A) and postinjection (B).

Downloaded for Anonymous User (n/a) at The University of Tennessee Health Science Center from ClinicalKey.com by Elsevier on August 17, 2021. For personal use only. No other uses without permission. Copyright ©2021. Elsevier Inc. All rights reserved. 34 Neumeister

Fig. 8. Before (A) and after (B) botulinum toxin laser scans of a patient who had previous amputations as a result of severe digital ischemia and necrosis. The patient had previous digital sympathectomies that did not improve the perfusion of the digits. The remaining digits had become ischemic and the patient underwent onabotulinum- toxinA injections to improve perfusion. The remaining digits were salvaged.

It may be that some patients have an immunologic 2. de Trafford JC, Lafferty K, Potter CE, et al. An epide- resistance to botulinum toxin; Borodic79 observed miological survey of Raynaud’s phenomenon. Eur J decreasing effectiveness of botulinum injections Vasc Surg 1988;2:167–70. after repeated treatment of facial rhytids. Earlier, 3. Department of Health and Human Services, Public Greene and colleagues80 described resistance to Health Service, National Institutes of Health, et al. botulinum toxin after repeated injections for pa- Questions and answers about Raynaud’s phenome- tients with torticollis. It is conceivable that some non. NIH; 2001. Publication No. 06–4911, Revised patients have localized resistance to the toxin June 2006. Available at: http://vickiehonea.com/ despite never having previous injections of the raynaudsRP_04.pdf. toxin. However, it is more likely that the patients 4. Garcia-Carrasco M, Jimenez-Hernandez M, with vasospastic disorders who do not respond Escarcega RO, et al. Treatment of Raynaud’s phe- as well as others may have alterations in the mech- nomenon. Autoimmun Rev 2008;8:62–8. anisms described earlier. 5. Bowling JC, Dowd PM. Raynaud’s disease. Lancet 2003;361:2078–80. SUMMARY 6. Fonseca C, Abraham D, Ponticos M. Neuronal regu- lators and vascular dysfunction in Raynaud’s phe- At present, botulinum injections offer a low-risk nomenon and systemic sclerosis. Curr Vasc method for the treatment of symptomatic Raynaud Pharmacol 2009;7(1):34–9. phenomenon. Although the mechanism has not 7. Ramien M, Brassard A. The challenge of scleroderma been fully elucidated, this finding provides patients ulcers. J Cutan Med Surg 2009;13(Suppl 1):S42–8. with another option for medical management 8. Nihtgyanova SI, Brough GM, Black CM, et al. Clin- before moving to sympathectomy. Repeated in- ical burden of digital vasculopathy in limited and jections may be required. diffuse cutaneous systemic sclerosis. Ann Rheum Dis 2008;67(1):120–3. ACKNOWLEDGMENTS 9. Morris JL, Jobling P, Gibbins IL. Differential inhibition by botulinum neurotoxin A of cotransmitters The author acknowledges the dedication of released from autonomic vasodilator neurons. Am Evyn Neumeister, BA, for her hours of work helping J Physiol Heart Circ Physiol 2001;28:H2124–32. to write and prepare this article. 10. Silveri F, De Angelis R, Poggi A, et al. Relative roles of endothelial cell damage and platelet activation in REFERENCES primary Raynaud’s phenomenon (RP) and RP sec- ondary to systemic sclerosis. Scand J Rheumatol 1. Bakst R, Merola JF, Franks AG, et al. Raynaud’s phe- 2001;30:290–6. nomenon: pathogenesis and management. J Am 11. Rosato E, Letizia C, Proietti M, et al. Plasma adreno- Acad Dermatol 2008;59(4):633–53. medullin and endothelin-1 levels are reduced and

Downloaded for Anonymous User (n/a) at The University of Tennessee Health Science Center from ClinicalKey.com by Elsevier on August 17, 2021. For personal use only. No other uses without permission. Copyright ©2021. Elsevier Inc. All rights reserved. Vasospastic Disorders of the Hand 35

Raynaud’s phenomenon improved by daily tadalafil 26. Bunker CB, Reavlety C, O’Shaughnessy DJ, et al. administration in male patients with systemic scle- Calcitonin gene-related peptide in treatment of se- rosis. J Biol Regul Homeost Agents 2009;23:23–9. vere peripheral vascular insufficiency in Raynaud’s 12. Durham PL, Cady R, Cady R. Regulation of calci- phenomenon. Lancet 1993;342:80–3. tonin gene-related peptide secretion from trigeminal 27. Generini S, Seibold JR, Matucci-Cerinic M. Estro- nerve cells by botulinum toxin type A: implications gens and neuropeptides in Raynaud’s phenomenon. for migraine therapy. Headache 2004 Jan;44(1): Rheum Dis Clin North Am 2005;31:177–86. 35–42 [discussion: 42–3]. 28. Chotani MA, Flavahan S, Mitra S, et al. Silent alpha 13. Edwards CM, Marshall JM, Pugh M. Cardiovascular (2C)-adrenergic receptors enable cold-induced responses evoked mild cool stimuli in primary vasoconstriction in cutaneous arteries. Am J Physiol Raynaud’s disease: the role of endothelin. Clin Sci Heart Circ Physiol 2000;278:H1075–83. 1999;96:577. 29. Bailey SR, Eid AH, Mitra S, et al. Rho kinase me- 14. Freedman RR, Girgis R, Mayes M. Endothelial and diates cold-induced construction of cutaneous ar- adrenergic dysfunction in Raynaud’s phenomenon teries: role of alpha2C-adrenoceptor translocation. and scleroderma. J Rheumatol 1999;26(11):2386–8. Circ Res 2004;94:1367–74. 15. Lazzerini PE, Capecchi PL, Bisogno S, et al. Homo- 30. Bailey SR, Mitra S, Flavahan S, et al. Reactive cysteine and Raynaud’s phenomenon: a review. oxygen species from smooth muscle mitochondria Autoimmun Rev 2010;9:181–7. initiate cold-induced constriction of cutaneous 16. Leppert J, Ringqvis A, Ahlner J. Seasonal variations arteries. Am J Physiol Heart Circ Physiol 2005;289: in cyclic GMP response on whole-body cooling in H243–50. women with primary Raynaud’s phenomenon. Clin 31. Furspan PB, Chatterjee S, Freedman RR. Increased Sci 1997;93:175. tyrosine phosphorylation mediates the cooling- 17. McMahon HT, Foran P, Dolly JO, et al. Tetanus toxin induced contraction and increase vascular reactivity and botulinum toxins type A and B inhibit glutamate, of Raynaud’s disease. Arthritis Rheum 2004;50: gamma-aminobutyric acid, aspartate, and met- 1578–85. enkephalin release from synaptosomes. Clues to 32. Furspan PB, Chatterjee S, Mayes MD, et al. Cooling- the locus of action. J Biol Chem 1992;267(30): induced contraction and protein tyrosine kinase 21338–43. activity of isolated arterioles in secondary Raynaud’s 18. Muller-Ladner U, Distler O, Ibba-Manneschi L, et al. phenomenon. Rheumatology 2005;44:488–94. Mechanisms of vascular damage in systemic scle- 33. Lau CS, McLaren M, Saniabadi A, et al. Increased rosis. Autoimmunity Oct 2009;42(7):587–95. whole blood platelet aggregation in patients with 19. Duprez DA. Role of the renin-angiotensin- Raynaud’s phenomenon with or without systemic aldosterone system in vascular remodeling and sclerosis. Scand J Rheumatol 1993;22:97–101. inflammation: a clinical review. J Hypertens 2006; 34. LeRoy EC. Systemic sclerosis: a vascular perspec- 24:983–91. tive. Rheum Dis Clin North Am 1996;22:675–94. 20. Kawaguchi Y, Takagi K, Hara M, et al. Angiotensin II 35. Flavahan NA, Flavahan S, Mitra S, et al. The vascul- in the lesional skin of systemic sclerosis patients opathy of Raynaud’s phenomenon and scleroderma. contributes to tissue fibrosis via angiotensin II type Rheum Dis Clin North Am 2003;29:275–91. 1 receptors. Arthritis Rheumatol 2004;50:216–26. 36. Henness S, Wigley FM. Current drug therapy for 21. Sgonc R, Gruschwitz MS, Boeck G, et al. Endothelial scleroderma and secondary Raynaud’s phenome- cell apoptosis in systemic sclerosis is induced by non: evidence-based review. Curr Opin Rheumatol antibody-dependent cell-mediated cytotoxicity via 2007;19(6):611–8. CD95. Arthritis Rheum 2000;43:2550–62. 37. Lambova SN, Muller-Ladner U. New lines in therapy 22. Flavahan NA, Vanhoutte PM. Endothelial cell of Raynaud’s phenomenon. Rheumatol Int 2009;29: signaling and endothelial dysfunction. Am J Hyper- 355–63. tens 1995;8(Suppl):28S–41S. 38. Neumeister MW, Chambers CB, Herron MS, et al. 23. Droste H, Wollersheim H, Reyenga J, et al. Vascular Botox therapy for ischemic digits. Plast Reconstr and humoral sympathetic nervous system reactively Surg 2009;124(1):191–201. during mental arithmetic in primary Raynaud’s 39. Van Beek AL, Lim PK, Gear AJL, et al. Management phenomenon. Int Angiol 1990;9:84–9. of vasospastic disorders with botulinum toxin A. 24. Bunker CB, Goldsmith PC, Leslie TA, et al. Calcitonin Plast Reconstr Surg 2007;119:217. gene-related peptide, endothelin-1, the cutaneous 40. Fregene A, Ditmars D, Siddiqui A. Botulinum toxin microvasculature and Raynaud’s phenomenon. Br type A: a treatment option for digital ischemia in J Dermatol 1996;134:399–406. patients with Raynaud’s phenomenon. J Hand Surg 25. Bunker CB, Terenghi G, Springall DR, et al. Am 2009;34(3):446–52. Deficiency of calcitonin gene-related peptide in 41. Koman LA, Smith BP, Smith TL. Vascular disorders. Raynaud’s phenomenon. Lancet 1990;336:1530–3. In: Green DP, Hotchkiss RN, Pederson WC, editors.

Downloaded for Anonymous User (n/a) at The University of Tennessee Health Science Center from ClinicalKey.com by Elsevier on August 17, 2021. For personal use only. No other uses without permission. Copyright ©2021. Elsevier Inc. All rights reserved. 36 Neumeister

Greens operative hand surgery. New York: Churchill 56. Burgen AS, Dickens F, Zatman LJ. The action of bot- Livingstone; 1999. p. 2254–302. ulinum toxin on the neuromuscular junction. 42. Baumhaekel M, Scheffler P, Boehm M. Use of J Physiol 1949;109(1–2):10–24. tadalafil in a patient with a secondary Raynaud’s 57. Setler P. Therapeutic use of botulinum toxins: back- phenomenon not responding to sildenafil. Microvasc ground and history. Clin J Pain 2002;18(6):S119–24. Res 2005;69(3):178–9. 58. Cheng CM, Chen JS, Patel RP. Unlabled uses of 43. Baak SW. Treatment of Raynaud’s phenomena with botulinum toxins: a review, part 1. Am J Health PDE-5 inhibitors, Cialis (tadalafil) in patients with Syst Pharm 2006;63(2):145–52. scleroderma and lupus. Arthritis Rheum 2005; 59. Aoki KR. Pharmacology and immunology of botulinum 52(Suppl 9):S169. toxin serotypes. J Neurol 2001;248(Suppl 1):3–10. 44. Cheung GT, Lau CS, Kumana CR. Phosphodies- 60. Neumeister MW. Botulinum toxin type A in the treat- terase 5-inhibitors relieve symptoms of severe ment of Raynaud’s phenomenon. J Hand Surg 2010; Raynaud’s phenomenon. Ann Rheum Dis 2004; 35A:2085–92. 64(Suppl 2):211. 61. Neumeister MW, Webb KN, Romanelli M. Mini- 45. Colglazier CL, Sutej PG, O’Rourke KS. Severe mally invasive treatment of Raynaud’s phenome- refractory fingertip ulcerations in a patient with non: the role of botulinum type A. Hand Clin scleroderma: successful treatment with sildenafil. 2014;30:17–34. J Rheumatol 2005;32(12):2440–2. 62. Sycha T, Graninger M, Auff E, et al. Botulinum toxin 46. Caglayan E, Huntgeburth M, Karasch T, et al. Phos- in the treatment of Raynaud’s phenomenon: a pilot phodiesterase type 5 inhibition is a novel therapeutic study. Eur J Clin Invest 2004;34:312–3. option in Raynaud’s disease. Arch Intern Med 2006; 63. Cui M, Khanijou S, Rubino J, et al. Subcutaneous 166:231–3. administration of botulinum toxin A reduces 47. Carlino G. Treatment of Raynaud’s phenomenon with formalin-induced pain. Pain 2004;107:125–33. tadalafil, a phosphodiesterase-5 inhibitor. Ann 64. Devor M, Wall PD. Cross-excitation in dorsal Rheum Dis 2005;64(Suppl 3):258. root ganglia of nerve-injured and intact rats. 48. Sandorfi N, Jimenez SA. Treatment of digital ulcera- J Neurophysiol 1990;64(6):1733–46. tion due to Raynaud’s phenomenon (RP) with long 65. Halperin JL, Cohen RA, Coffman JD. Digital vasodi- acting phosphodiesterase 5 (PDE5) inhibitors: tada- latation during mental stress in patients with Ray- lafil. Ann Rheum Dis 2007;66(Suppl 2):223. naud’s disease. Cardiovasc Res 1983;17:671–7. 49. Kumar N, Griffiths B, Allen J. Thermographic and 66. Colhado OC, Boeing M, Ortega LB. Botulinum toxin symptomatic effects of a single dose of sildenafil in pain treatment. Rev Bras Anestesiol 2009;59(3): citrate on Raynaud’s phenomenon in patients with 366–81. systemic sclerosis: a potential treatment. 67. Martinez RM, Saponaro A, Dragagna G, et al. Cuta- J Rheumatol 2006;33(9):1918–9. neous circulation in Raynaud’s phenomenon during 50. Kamata Y, Kamimura T, Iwamoto M, et al. Compara- emotional stress: a morphological and functional ble effects of sildenafil citrate and alprostadil on study using capillaroscopy and laser-Doppler. Int severe Raynaud’s phenomenon in a patient with Angiol 1992;11:316–20. systemic sclerosis. Clin Exp Dermatol 2005;30: 68. Fathi M, Fathi H, Mazloumi M, et al. Preventive effect 435–56. of botulinum toxin A in microanastomotic throm- 51. Lin CS, Lin G, Xin ZC, et al. Expression, distribution, bosis: a rabbit model. J Plast Reconstr Aesthet and regulation of phosphodiesterase 5. Curr Pharm Surg 2010;63(10):e720–4. Des 2006;12:3439–57. 69. Janz BA, Thomas PR, Fanua SP, et al. Prevention of 52. Pakozdi A, Howell K, Black CM, et al. Addition of the anastomotic thrombosis by botulinum toxin B after short term phosphodiesterase-5 inhibitor sildenafil acute injury in a rat model. J Hand Surg Am 2011; to iloprost therapy for scleroderma digital vasculop- 36(1):1585–91. athy. Arthritis Rheum 2007;56(Suppl):825. 70. Park BY, Kim HK, Kim WS, et al. The effect of botu- 53. Schiopu E, Hsu VM, Impens AJ, et al. Controlled trial linum toxin B pretreatment to the blood flow in the of tadalafil in Raynaud’s phenomenon (RP) second- microvascular anastomosis. Ann Plast Surg 2014; ary to systemic sclerosis. Arthritis Rheum 2006; 72(2):214–9. 54(Suppl 2):505. 71. Murakami E, Iwata H, Imaizumi M, et al. Prevention 54. Zamiri B, Koman AL, Smith BP, et al. Double-blind, of arterial graft spasm by botulinum toxin: an in- placebo-controlled trail of sildenafil for the manage- vitro experiment. Interact Cardiovasc Thorac Surg ment of primary Raynaud’s phenomenon. Ann 2009;9(3):395–8. Rheum Dis 2004;63(Suppl3):342. 72. Clemens MW, Higgins JP, Wilgis EF. Prevention of 55. Snipe PT, Sommer H. Studies on Botulinus toxin: 3. anastomotic thrombosis by botulinum toxin-A in an Acid precipitation of Botulinus toxin. J Infect Dis animal model. Plast Reconstr Surg 2009;123(1): 1928;43(2):152–60. 64–70.

Downloaded for Anonymous User (n/a) at The University of Tennessee Health Science Center from ClinicalKey.com by Elsevier on August 17, 2021. For personal use only. No other uses without permission. Copyright ©2021. Elsevier Inc. All rights reserved. Vasospastic Disorders of the Hand 37

73. Upton J, Garcia J, Liao E. Botox to the rescue. Plast 77. Schweizer DF, Schweizer R, Zhang S, et al. Botuli- Reconstr Surg 2009;123(1):38e. num toxin A and B raise blood flow and increase sur- 74. Kim YS, Roh TS, Lee WJ, et al. The effect of botuli- vival of critically ischemic skin flaps. J Surg Res num toxin A on skin flap survival in rats. Wound 2013;184:1205–13. Repair Regen 2009;17(3):411–7. 78. Stone AV, Koman LA, Callahan MF, et al. The effect 75. Arnold PB, Merritt W, Rodeheaver GT, et al. Effects of botulinum neurotoxin A on blood flow in rats: a of perivascular botulinum toxin-A application on potential mechanism for the treatment of Raynaud’s vascular smooth muscle and flap viability in the phenomenon. J Hand Surg 2012;37(4):795–802. rat. Ann Plast Surg 2009;62(5):463–7. 79. Borodic G. Immunologic resistance after repeated 76. Arnold PB, Fang T, Songcharoen SJ, et al. Inflamma- botulinum toxin A injections for facial rhytids. Oph- tory response and survival of pedicled abdominal thal Plast Reconstr Surg 2006;6(22):239. flaps in a rat model after perivascular application 80. Greene P, Fahn S, Diamond B. Development of of botulinum toxin type A. Plast Reconstr Surg resistance to botulinum toxin type A in patients 2014;133(4):491e–8e. with torticollis. Mov Disord 1994;9:213.

Downloaded for Anonymous User (n/a) at The University of Tennessee Health Science Center from ClinicalKey.com by Elsevier on August 17, 2021. For personal use only. No other uses without permission. Copyright ©2021. Elsevier Inc. All rights reserved. Curr Rev Musculoskelet Med (2017) 10:567–574 https://doi.org/10.1007/s12178-017-9451-2

QUALITY AND COST CONTROL IN TJA (B WADDELL, SECTION EDITOR)

Outpatient Total Joint Arthroplasty

Jack M. Bert1,2 & Jessica Hooper3 & Sam Moen1

Published online: 24 October 2017 # Springer Science+Business Media, LLC 2017

Abstract Introduction Purpose of Review Outpatient total joint arthroplasty (OTJA) allows for a safe, cost effective pathway for appropriately se- Outpatient total joint arthroplasty (OTJA) is a means of provid- lected patients. With current pressures on arthroplasty surgeons ing safe, cost effective care for select patients indicated for and their associated institutions to reduce costs per episode of unicompartmental knee arthroplasty (UKA), total knee care, it is important to define the steps and challenges associ- arthroplasty (TKA), or total hip arthroplasty (THA). ated with establishing an outpatient arthroplasty program. Establishing a successful OTJA program involves meticulous Recent Findings Several studies have outlined techniques of attention to detail and anticipation of potential complications. In selecting patients suitable for this type of postoperative path- addition to preparing the surgery center, surgeons must also way. With emerging concerns about patients who undergo establish procedure to optimize anesthetic technique, pain man- outpatient arthroplasty being at increased risk of medical com- agement, blood management, rehab and physiotherapy, and plications, which may lessen projected cost savings, it is im- patient safety [1]. Despite the immense amount of planning portant to identify value-based strategies to optimize patient required, the rate of arthroplasty procedures done in the outpa- recovery after OTJA. tient setting is increasing. Between 2012 to 2015, there was a Summary This article reviews digital techniques for patient 47% increase in elective OTJA, and it is expected that there will selection and data collection, operating room efficiency sys- be a 77% growth in OTJA over the next 10 years, with inpatient tems, and provides a summary of methods to build and main- TJA growing only 3% during the same period (Fig. 1)[2]. tain value in outpatient total joint replacement within the In 2014, Medicare paid $50,000 per TJA hospitalization, framework of bundled payment reimbursement. totaling approximately $7 billion paid out that year [3]. In 2011, the Center for Medicare and Medicaid Services (CMS) Keywords Outpatient total joint arthroplasty . Patient introduced the Bundled Payment for Care Improvement (BPCI) stratification . Episode of care management with the goal of establishing a payment model that would lead to “higher quality, more coordinated care at a lower cost to This article is part of the Topical Collection on Quality and Cost Control ” in TJA Medicare [4]. This model incentivizes improving quality while cutting costs in order to improve the value generated per unit * Jack M. Bert arthroplasty procedure. In practice, this payment model pro- [email protected] motes two primary goals: first, to reduce nonessential operating * Sam Moen room (OR) and hospital services and minimize adverse events [email protected] requiring increased length of stay, readmission, and/or dis- charge to inpatient rehab units, and second, to better coordinate 1 730 N 4th Street Unit 113, Minneapolis, MN 55401, USA care for arthroplasty patients between the inpatient and outpa- 2 Minnesota Bone and Joint Specialists, Ltd, 2025 Woodlane Drive, tient settings. By streamlining the surgical procedure and elim- Woodbury, MN 55125, USA inating the postoperative stay, OTJA accomplishes both goals. 3 NYU Langone Orthopedic Hospital, 301 E 17th St #1402, New Bundling orthopedic knee and hip procedures requires a York, NY 10003, USA different approach from traditional fee-for-service practice. 568 Curr Rev Musculoskelet Med (2017) 10:567–574

Fig. 1 Primary hip and knee replacement growth across settings [2]

Bundles include all areas of the surgical process beginning the location of the surgery, it is the responsibility of the sur- with the surgical consult and in most cases ending 90 days geon to take all measures necessary to keep costs down with- after the surgical procedure. Paramount to episodic bundling out compromising patient outcomes or experiences. best practice is the preoperative clinical and social risk screen- ing to identify risks for high cost and low quality and manag- ing these risk factors prior to surgery. Understanding the pa- Insurance Contracting tient’s risks is a proactive way to effectively manage and elim- inate all risk factors reducing cost and increasing quality of the Patient and ambulatory surgery center (ASC) insurance patient’s surgical outcome. contracting is critical to make certain that the insurance pre- A successful bundled program for total joint procedures approval process identifies any facility-fee reimbursement is- requires a cultural shift in the way a case is managed including sues as well as non-covered patient costs. It is not uncommon predictive risk assessments/scoring and shared decision-mak- for some insurers to consider TJA performed at an ASC an ing. Figure 2 outlines the components of a bundled payment “out of network” procedure, as opposed to the “in-network” program. TJA performed the local hospital. It is critical for the ASC to Organizations that participate in bundled payments risk sig- obtain credentialing for OTJA and do a thorough pro-forma nificant financial loss. This is because the bundle assumes that based upon payer mix, meet with the payers, and negotiate a care for all patients will be covered by the single bundle pay- bundled payment rate for each TJA procedure. ment. However, a small subset of patients, termed “avalanche For Medicare patients, insurance contracting can discour- events” or “bundle busters,” can have increased cost of care due age OTJA. Currently, CMS will reimburse UKA for same day to multiple-comorbidities or complications resulting in hospital discharge and allows overnight stays or admissions under re-admissions and this can cost between $100,000 to $150,000 “observation” class, for less than 24 h without penalty. per admission [5]. These patients, when mismanaged, fall out- Conversely, many hospitals are charged an early discharge side the organization’s process and capabilities. Organizations penalty for outpatient TKA. Identifying these situations ahead looking to become “best in class” have the opportunity to re- of time can help minimize undue financial burden on the pa- duce the risk factors for every patient prior to surgery, which tient and is an essential component of a comprehensive will result in the most optimized and low-risk patients. These screening program for OTJA. results in a reduced expense to the bundled payment with high- quality outcomes and increased patient satisfaction. In our ex- perience, identification of the patients’ clinical and social risk factors has been critically important in determining the most Patient Selection and Screening important risk mitigation interventions prior to OTJA. Although the future of the CMS BPCI initiative for TJA is Establishing patient selection criteria is the most critical aspect uncertain, there is bipartisan support in congress for the cost of OTJA, with several articles written in the past year savings promoted by the Comprehensive Care for Joint outlining the necessary criteria to avoid complications. Sibia Replacement (CJR) program [6]. As further evidence that et al. reviewed a series of 381 TKAs and found that older age, the ways in which we provide care to arthroplasty patients female gender, ASA scores of III/IV, history of atrial fibrilla- are changing, the Advisory Panel on Outpatient Payment tion, and prior TKA were associated with increased hospital unanimously recommended that CMS move the CPD code costs and increased length of stays [8]. Berger et al. reported for TKA (27447) from inpatient-only status to hospital outpa- on 5373 TJAs between 2004 and 2013 and confirmed that tient department (HOPD) status in August 2016 [7]. No matter patients who received medical clearance and the coordinated Curr Rev Musculoskelet Med (2017) 10:567–574 569

Fig. 2 Organizing the episode of care

efforts of a dedicated postoperative caregiver and a discharge reported questionnaire for screening and measuring general planner have better outcomes after TJA [9]. anxiety disorder, is a useful tool and can be used in concert The only way to ensure a consistent and manageable epi- with other tools for screening overall health, such as the re- sode of care is to mandate that each patient has a detailed risk vised cardiac risk index (RCRI). For example, a patient may assessment prior to their surgical event. Risk assessments give have three identified risk factors: the surgical team accurate, updated information on the pa- tient’s risks of perioperative complications. This information & High body mass index (BMI) also helps the surgeon determine whether a patient can safely & Symptomatic chronic heart failure proceed with OTJA; they are better candidates for inpatient & Anxiety TJA, or further medical evaluation and optimization are nec- essary prior to surgery. These risk factors need to be properly managed and tracked Unlike traditional medical risk stratification for TJA, it is im- until they are clinically capable of proceeding with surgery. portant to evaluate anxiety levels and social support for patients Possible interventions for this patient may be setting up an being considered for OTJA. Even patients without diagnosed appointment with a dietician, working with their primary care anxiety disorder are often concerned about postoperative pain, physician and cardiologist to manage congestive heart failure being able to care for themselves, and mobility, especially if they (CHF), and setting them up with psychotherapy prior to their live alone. A study of the effect of psychological support on visit. Risk calculations and historical data tell us the obese patients indicated for primary TJA published by Tristano et al. patients have higher risk for readmissions and complications demonstrated that patients who received psychological support [11]. If these issues are not managed, this patient could be- had a lower incidence of anxiety and depression than those who come one of the “bundle busters,” and experience complica- did not [10]. Additionally, THA patients who received psycho- tions and/or costly readmissions that may have been avoidable logical support reached their physical therapy goals 1.2 days with more thorough preoperative optimization. sooner than those who did not receive any psychological support Making sure the patient is healthy, properly managed, and [10]. Patients being considered for OTJA, who are required to ready for surgery is the difference between a successful and manage their own pain control, physical therapy, and self-care, unsuccessful bundled program. Organizations may be able to are even more likely to benefit from a strong psychological sup- manage a low volume of bundled patients without detailed port. Identifying patients who already have strong social support protocols, process, and technologies. However, once the vol- vs. those who do not is thus extremely beneficial. ume increases, it quickly becomes more difficult for providers A detailed clinical and social assessment such as the to keep up with all patients and understand how each of them General Anxiety Disorder-7 (GAD-7), which is a self- is responding to their medical management. Reactive 570 Curr Rev Musculoskelet Med (2017) 10:567–574 management, responding after a complication or readmission same-day discharge [10]. They found no differences in BMI, has occurred, is not advisable when caring for large groups of average age, body weight, or medical comorbidities between patients; anticipatory management enables adherence to the patients who underwent OTJA and those who required inpa- OTJA strict protocols. tient stay [13]. They also found that for patients undergoing Data collection and outcome measures are very important TKA, compared to UKA, history of myocardial infarction and can be accomplished with a software program (Fig. 3). within 1 year, BMI > 40, and greater than three medical co- Tracking volume, surgical outcomes, and patient progress morbidities were most strongly associated with increased risk help surgeons identify successful practices and areas for im- of readmission [14]. provement. A “total joint coordinator,”“nurse manager,” or In their review of 1012 patients undergoing elective TJA, “case manager” must be assigned to every patient to make Courtney et al. found that chronic obstructive pulmonary dis- certain that the surgeons’ and patients’ needs are being met. ease (COPD), coronary artery disease (CAD), congestive Additionally, this individual ensures that the necessary staff heart failure (CHF), and cirrhosis were identified as indepen- and service line resources have been coordinated successfully. dent risk factors for late (> 24 h) postoperative complications This individual must do a patient home assessment, which is [15]. As such, the authors do not recommend OTJA for pa- part of the software program (Fig. 3) to make certain that the tients with any of these comorbidities. They proposed a 6- home caregiver is competent to tend to the needs of the patient point risk assessment model (AUC = 0.738) to help determine in the first 48 to 72 h postoperatively. This person also helps apatient’s candidacy for short stay TJA; patients without any coordinate preoperative multidisciplinary patient education of the aforementioned risk factors have a 3.1% probability of and preoperative physical therapy, to ensure patients in the developing late complications, while the presence of one risk best possible condition before surgery [12]. Home physical factor elevates the patient’sriskto10%[15]. therapy can be instituted by video learning tools or utilizing These studies, rather than providing steadfast indication a mobile physical therapist making home visits. and contraindication criteria for patients undergoing outpa- tient TJA, serve to underscore the importance of thorough preoperative medical evaluation. Also, established protocols Cautions—Identifying Patients Who Are Not for therapy and pain control are essential to help patients be Candidates for OTJA successfully discharged from ASCs; disorganized postopera- tive treatment can derail OTJA for even the most medically Published literature is mixed on medical comorbidities associ- and psychologically optimized patient. ated with perioperative complications and readmissions fol- lowing OTJA. In their retrospective study of 110 patients who underwent OTJA at a center accustomed to rapid rehab ASC Facility and Staff Preparation protocols and minimally invasive techniques, Berger et al. found that inadequate pain control and completing surgery late While careful patient selection is essential to minimize risks in the day were most strongly associated with preventing and ensure optimal candidates for OTJA surgery, ambulatory

Fig. 3 Data collection and outcome measures Curr Rev Musculoskelet Med (2017) 10:567–574 571 care facilities also pose new challenges and considerations for areas of the outpatient facility, such as central processing. surgeons. These include the following: Digital solutions can also support the streamlining and stan- dardization of instrument trays to help decrease instrumenta- & Training staff to be proficient on multiple procedures as tion setup time and reduce costs associated with sterilizing well as each surgeon’s exact procedure preferences and unused instrumentation. instrument sequencing & Minimizing instruments to be sterilized for time and cost savings Interactive Training & Timely turnover of operating rooms to maximize OR uti- lization and patient throughput With the ability to deliver interactive training modules that rapidly onboard staff and reinforce protocols and procedures, In an ambulatory care facility, outpatient total joint replace- while also enabling customized training for surgeon-specific ment requires an efficient and streamlined OR and central preferences and instrument sequencing, digital solutions build processing. Two areas, staff training and instrumentation, sig- OR staff competency. By digitizing surgeon preference cards nificantly impact the success of outpatient total joint replace- and procedure instrument sequencing, these assets can be ment procedures and the ability to reduce costs. At the same placed online for staff to review and facilitate an interactive time, the use of clinically proven hip and knee implants can experience that leverages gaming technology. This allows help reduce the risk of complications. training and reinforcement, to take place anywhere, anytime, While not unique to an ambulatory care setting, staff train- at staff convenience. Flexible solutions that support on-site ing in a high-volume, time-sensitive environment is vital to training as well as remote access to training modules and achieving the benefits offered by OTJA. The ability to train can support a breadth of content for procedure types and pro- staff on a surgeon’s specific procedure and instrument se- tocols in one place can empower staff and increase confidence quencing can affect procedure duration, and staff trained on in an outpatient environment. The benefits of interactive train- a surgeon’s preferred patient setup and room layout can facil- ing platforms for outpatient total joint replacement include the itate faster OR turnover. Enabling easy and fast access to in- following: structions for robotic or other complex equipment also sup- ports procedure efficiency and speed. & Decreased OR time Not only is staff competence improved with easy access to & Decreased time to onboard staff training and self-testing prior to procedures, ensuring profi- & Reinforce protocols and procedure preferences prior to a ciency in the OR also saves time in an ambulatory environ- case ment. When systems and programs are put in place to support & Improve staff competence the needs of each surgeon, the procedures performed and in- & Increase surgeon confidence in staff strument sequencing, surgeon’s confidence in outpatient OR staff increases. The ability to reduce costs is a significant driver for the increase in OTJA. Streamlining the OR via the standardization Streamlined Instrumentation of instrument trays can decrease setup time and reduce the costs associated with sterilizing instruments that go unused. Instrument utilization identification is used to streamline and With the ability to only pull the instruments, a surgeon uses, standardize instrument trays, which can help decrease setup with back-up instruments sterilized and ready but unopened, time and reduce the costs associated with sterilizing unused outpatient total joint replacement procedures can save the instruments. By “right-sizing” trays based upon surgeon and costs of sterilizing unused instruments while also reducing procedure requirements, streamlining trays for outpatient total procedure time. joint replacement also helps ensure that shelf space is opti- mized for the reduced footprint in ambulatory facilities. Technology platforms designed to streamline and standardize Technology-Enabled Improvements Facilitate instrument trays can as follows: Outpatient Total Joint Replacement & Help decrease setup time Advancements in technology-enabled solutions can help en- & Reduce costs associated with sterilizing unused instruments sure outpatient OR staff proficiency and facilitate efficiency in & Maximize shelf space OTJA procedures. These digital solutions leverage the conve- & Provide a visual platform for tray assembly and special nience of self-guided, personalized training for staff prior to instructions procedures and can also be deployed to help training in other & Document and validate proper tray inspection 572 Curr Rev Musculoskelet Med (2017) 10:567–574

One example of a cloud-based technology to improve out- mobilization necessary for discharge. Utilizing these tech- patient OR efficiency and enhance patient care is a surgeon niques allows the patient to be ambulatory within a few hours specific suite called Virtual BackTable (S2 Interactive, Smith after surgery without significant pain. Cold therapy has been and Nephew, Andover, MA) that combines digital interactive shown to be beneficial in the early postoperative period as training and instrument utilization identification to help im- well. prove staff competency and streamline trays to reduce instru- It is important to avoid significant blood loss and bleeding mentation and sterilization costs. at the time of surgical intervention, both in the interest of In one pilot study, using S2 technology resulted in a 60% preventing symptomatic anemia, and also because blood is a reduction in instrument tray processing costs, with an estimat- noxious stimulus and massive hemarthrosis can cause signif- ed savings of $83,000 per year for just one tray [15]. At the icant pain delaying progress with physical therapy [19]. Some same time, the technology also facilitates improved workflow OTJA surgeons avoid using a tourniquet completely during and support for compliance and quality protocols. TKA and keep a tourniquet in place to inflate only if any In OTJA procedures, the costs of instrument processing untoward bleeding ensues during the procedure [20]. If using and implant waste can quickly add up and impact the very a tourniquet for a TKA, it is important to release the tourniquet cost-saving benefits both surgeons and patients seek to realize prior to closure and coagulate all major bleeders prior to soft through OTJA. In addition, the inefficiencies associated with tissue injection of using one of the anesthetic “cocktails” or a insufficiently trained staff and large instrument inventories bupivacaine liposomal injection. Regional anesthesia, by pre- can also impact the potential cost benefits that can be achieved cipitating a hypotensive state, also helps decrease blood loss, in an ambulatory environment. Regardless of the solution, especially from cut bone surfaces in UKA and TKA [21, 22]. technology can play an important role in supporting the real- The use of perioperative tranexamic acid (TXA), which has ization of the time and cost saving benefits that outpatient total been used successful in primary and revision TJA without a joint replacement can offer. corresponding increase in the incidence of thromboembolic events [23, 24], is critically important to minimize blood loss. Any unexpected blood loss will slow patients’ recovery after Intraoperative Considerations: Minimizing Adverse surgery and can lead to the conversion of an OTJA case to an Events inpatient stay.

Multimodal pain management pathways have become the standard of care for TJA in the inpatient and outpatient set- Outcomes tings [16, 17]. It is of particular importance for patients under- going outpatient TJA that an effective anesthesia program be In the last 10 years, many arthroplasty surgeons have made the in place to control pain, minimize adverse reactions to anes- move toward performing OTJA for patients without serious thesia, such as nausea, and allow rapid mobilization after sur- medical or psychiatric comorbidities. Published literature has gery. Ideally, pain and anesthesia pathways should be shown that OTJA can be safely performed without negatively established by a team consisting of anesthesiologists, sur- affecting patient outcomes or patient perceptions of the quality geons, and pain management physicians and should combine of their care. oral medications and local anesthetics. These protocols would With the development of fast-track arthroplasty programs, reduce the need for oral medications and prevent peripheral expedited recovery programs, such as OTJA, have been sensitization by neurogenic blockade [11]. shown to improve efficiency while maintaining low-revision Adductor canal blocks are the procedure of choice for pa- rates, improved quality of life and functionality, and high- tients undergoing either TKA or UKA, and short-acting spinal patient satisfaction scores [25]. OTJA programs have demon- blocks are effective for THA. Short-acting spinal anesthetic strated the ability to maintain the positive outcomes of the fast- agents often consist only of local anesthetics, such as lidocaine track programs without undue increase in risk to patients. or ropivacaine, which helps minimize the potential nausea, In their study comparing rates of perioperative complica- sedation, and pruritus, that occur with longer-acting anes- tions between patients who underwent inpatient and outpatient thetics containing opioids [18]. Injection of the capsular and UKA, Bovonratwet et al. reported no significant differences in soft tissue structures by the surgeon may include one of vari- complication rate or 30-day readmission rate between the ous reported “cocktails,” often consisting of a combination of matched patient groups [26]. Dorr et al. reported on a series narcotics, local anesthetics, and non-steroidal anti-inflamma- of patients offered outpatient THA and found that, though tory drugs (NSAIDs) and/or the utilization of a liposomal outpatient surgery offered no objective physical benefits, such bupivicaine injection. Another important aspect of periopera- as accelerated functional recovery, it did not result in any tive care is preventing dehydration, which can exacerbate pos- medical complications or readmissions [27]. Their results tural hypotension and therefore delay postoperative demonstrated that THA could safely be performed for a select Curr Rev Musculoskelet Med (2017) 10:567–574 573 group of patients and that these patients were also willing to avoid unnecessary medical interventions. Though not all pa- accept the responsibilities associated with outpatient surgery, tients will be candidates for OTJA, the future of total joint underscoring the importance of patient engagement in care to arthroplasty will involve a movement to the outpatient setting a successful outcome. Six weeks after surgery, 87% of patients for the average patient. By shifting the setting of postoperative reported that going home the day of surgery increased their care from hospitals to patients’ homes, costs to patients are confidence in their hip replacement [27]. less [32] and help patients become more engaged in their care. Aynardi et al. compared outcomes and costs between pa- Building a successful OTJA endeavor requires the input of a tients who underwent inpatient THA and those who had out- multidisciplinary team and allows surgeons to affect the way patient THA. All surgeries were performed by the same sur- comprehensive care is delivered to arthroplasty patients. geon through the same approach, and there were no significant differences between the two groups with respect to complica- Compliance with Ethical Standards tions or blood loss. Importantly, the authors reported that the non-itemized bill charged to the patient or third-party payer Conflict of Interest All authors declare that they have no conflict of interest. was nearly $7000 less for the outpatient cases [28]. Similarly, a case control study conducted by Huang et al. comparing Human and Animal Rights and Informed Consent This article does costs for inpatient and outpatient TKA demonstrated that out- not contain any studies with human or animal subjects performed by any patient TKA, on average, resulted in 30% cost savings for the of the authors. hospital per episode of care without a significant increase in complications or readmissions [29]. These results demonstrate that, so long as the infrastructure is in place for outpatient References arthroplasty and patients are appropriately selected, OTJA can be considered a means of cost savings for hospitals and 1. Kehlet H, Søballe K. Fast-track hip and knee replacement — what – can thus increase value of care within a bundled-payment are the issues? Acta Orthop. 2010;81(3):271 2. 2. Impact of Change® v16.0; HCUP National Inpatient Sample (NIS). reimbursement model. Healthcare cost and utilization project (HCUP). 2013. Agency for Healthcare Research and Quality, Rockville, MD; OptumInsight, 2014; The following 2014 CMS Limited Data Sets (LDS): Conclusions Carrier, Denominator, Home Health Agency, Hospice, Outpatient, Skilled Nursing Facility; The Nielsen Company, LLC, 2016; Sg2 Analysis, 2016. OTJA is economically and practically feasible and yields ex- 3. Courtney P, Boniello A, Berger R. Complications following outpa- cellent outcomes in carefully selected patients. A well- tient total joint arthroplasty: an analysis of a national database. J designed OTJA program provides safe surgical care with Arthroplast. 2017;32:1426–30. equivalent patient-reported outcomes, such as anxiety, depres- 4. Iorio R. Strategies and tactics for successful implementation of “ bundled payments: bundled payment for care improvement at a sion, satisfaction, and pain, compared to standard fast track large, urban, Academic Medical Center. J Arthroplast. patients” that require a 1-day hospitalization [30, 31]. 2015;30(3):349–50. Furthermore, OTJA programs generate cost savings by im- 5. Ellimoottil C, Ryan A, Hou H, Dupree J, Hallstrom B, Miller D. proving OR efficiency via surgeon-specific instrumentation The new bundled payment program for joint replacement may un- fairly penalize hospitals that treat patients with medical comorbid- and online training tools and improving post-anesthesia care ities. Health Aff (Millwood). 2016;35(9):1651–7. and pain control. Postoperative home physical therapy train- 6. Caffrey M. The future of the comprehensive care for joint replace- ing, either by a therapist or with modules accessible by the ment program. Am J Manag Care. 2017. patient’s cell phone, facilitate safe, timely pre-habilitation and 7. Gollogly S Safety and efficacy of outpatient arthroplasty. Advisory rehabilitation, and obviate the need for patient transfer to a panel on outpatient payments. Center for Medicare and Medicaid services report. Woodlawn, Maryland, August 22, 2016. skilled nursing or extended care facility. It is critical to under- 8. Sibia US, King PJ, MacDonald JH. Who is not a candidate for a 1- stand that patient stratification and case management are of day hospital-based total knee arthroplasty. J Arthroplast. paramount importance in the success of the procedure, as well 2017;32(1):16–9. as controlling the total cost of the episode of care. Choosing 9. Berger RA, Cross MB, Sanders S. Outpatient hip and knee replace- “ ” “ ment: the experience from the first 15 years. Instr Course Lect. the wrong patient can cause an avalanche case or bundle – ” 2016;65:547 51. buster, which reaffirms that, without appropriate patient se- 10. Tristaino V,Lantieri F, Tornago S, Gramazio M, Carriere E, Camera lection and monitoring, the remainder of the care pathway will A. Effectiveness of psychological support in patients undergoing be doomed to failure. primary total hip or knee arthroplasty: a controlled cohort study. J Rigorous medical screening and optimization, along with Orthop Traumatol. 2016;17(2):137–47. multi-specialty care programs to control pain, limit blood loss, 11. Argenson J-NA, Husted H, Lombardi A, Booth RE, Thienpont E. Global forum: an international perspective on outpatient surgical minimize adverse reactions to anesthesia, and mobilize more procedures for adult hip and knee reconstruction. J Bone Joint quickly, help patients recover from surgery more quickly and Surg. 2016;98(13):e55. 574 Curr Rev Musculoskelet Med (2017) 10:567–574

12. Reinke C, Kelz R, Zubizarreta J, Lanyu M, Saynisch P, Kyle F, et al. 23. Wind TC, Barfield WR, Moskal JT. The effect of tranexamic acid Obesity and readmission in elderly surgical patients. Surgery. on transfusion rate in primary total hip arthroplasty. J Arthroplast. 2012;152(3):355–62. 2014;29(2):387–9. 13. Berger RA, Kusuma SK, Sanders SA, Thill ES, Sporer SM. The 24. Yang ZG, Chen WP, Wu LD. Effectiveness and safety of feasibility and perioperative complications of outpatient knee tranexamic acid in reducing blood loss in total knee arthroplasty. Arthroplasty. Clin Orthop Relat Res. 2009;467(6):1443–9. J Bone Joint Surg. 2012;94(13):1153–9. 14. Courtney PM, Rozell JC, Melnic CM, Lee G-C. Who should not 25. Winther SB, Foss OA, Wik TS, Davis SP, Engdal M, Jessen V,et al. undergo short stay hip and knee arthroplasty? Risk factors associ- 1-year follow-up of 920 hip and knee arthroplasty patients after ated with major medical complications following primary total joint implementing fast-track: good outcomes in a Norwegian university arthroplasty. J Arthroplast. 2015;30(9):1–4. hospital. Acta Orthop. 2015;86(1):78–85. 15. Hensell MG, Pins J. Virtual BackTable. Reducing cost of on- 26. Bovonratwet P, Ondeck NT, Tyagi V, Nelson SJ, Rubin LE, Grauer boarding. AORN Annual Meeting. Denver, Colorado, March 7– JN. Outpatient and inpatient unicompartmental knee arthroplasty 11, 2015. procedures have similar short-term complication profiles. J – 16. Berend ME, Berend KR, Lombardi AV. Advances in pain manage- Arthroplast. 2017;S0883-5403(17):30443 6. ment. Bone Joint J. 2014;96(11):7–9. 27. Dorr LD, Thomas DJ, Zhu J, Dastane M, Chao L, Long WT. Outpatient total hip arthroplasty. J Arthroplast. 2010;25(4):501–6. 17. Parvizi J, Miller AG, Gandhi K. Multimodal pain management after 28. Aynardi M, Post Z, Ong A, Orozco F, Sukin DC. Outpatient surgery total joint arthroplasty. J Bone Joint Journal. 2011;93(11):1075–84. as a means of cost reduction in Total hip Arthroplasty: a case- 18. Lombardi A, Barrington JW, Berend KR, Berend ME, Dorr LD, control study. HSS J. 2014;10(3):252–5. Hamilton W, et al. Outpatient arthroplasty is here now. Instr Course – 29. Huang A, Ryu J-J, Dervin G. Cost savings of outpatient versus Lect. 2016;65:531 46. standard inpatient total knee arthroplasty. Can J Surg. 2017;60(1): 19. Liu D, Dan M, Martinez Martos S, Beller E. Blood management 57. strategies in total knee arthroplasty. Knee Surg Rel Res. 2016;26(3): 30. Hoorntje A, Koenraadt KLM, Boevé MG, van Geenen RCI. – 179 87. Outpatient unicompartmental knee arthroplasty: who is afraid of 20. Aglietti P, Baldini A, Vena LM, et al. Effect of tourniquet use on outpatient surgery. Knee Surg Sports Tramatol Arthrosc. activation of coagulation in total knee replacement. Clin Orthop Rel 2017;25(3):759–66. Res. 2000;371:169–77. 31. Lovecchio, et al. Is outpatient arthroplasty as safe as fast-track in- 21. Sculco TP,Baldini A, Keating EM. Blood management in total joint patient arthroplasty? A propensity score matched analysis. J arthroplasty. Instr Course Lec. 2005;54:51–66. Arthroplast. 2016;31:197–201. 22. Hooper J, Schwarzkopf R. Additional tools to prevent blood loss in 32. Kaplan RS, Porter ME. How to solve the cost crisis in health care. total joint arthroplasty. Tech Orthop. 2017;32(1):34–40. Harv Bus Rev. 2011;89(9):46–52. Feature Article

Obesity, Young Age, Patellofemoral Disease, and Anterior Knee Pain: Identifying the Unicondylar Arthroplasty Patient in the United States

KEITH R. BEREND, MD; ADOLPH V. LOMBARDI, jR, MD; JOANNE B. ADAMS, BFA

abstract

Standard surgical criteria for unicondylar knee arthroplasty exclude patients in patients in the United States who do not who weigh >82 kg (187 lbs), patients aged <60 years, and patients with more meet the standard inclusion criteria. Specifi- than minimal erosive changes in the patellofemoral articulation, indicated cally, the current study investigates the role of clinically by anterior knee pain. In April 2004, the US Food and Drug Admin- obesity, young age, patellofemoral arthritis, istration approved the Oxford phase 3 unicondylar prosthesis for implantation and isolated medial knee pain on the early during unicompartmental knee arthroplasty. Since this broader indication was outcomes and failures of this device. adopted, the implant has proven to be a bone- and ligament- conservative yet long-lasting intervention for anteromedial arthritis of the knee. The purpose MATERIALS AND METHODS of this study is to investigate the early outcomes of unicompartmental medial Demographics. Between July 2004 knee arthroplasty using the Oxford prosthesis in US patients who do not meet and December 2006, 318 unicompartmen- the standard inclusion criteria. Specifically, the current study investigates the tal knee arthroplasties were performed in impact of obesity, young age, patellofemoral arthritis, and isolated medial knee 270 patients. Nineteen patients underwent pain on the early outcomes of this device. In this large series of 318 unicom- simultaneous bilateral unicompartmental partmental knee arthroplasties performed in 270 patients, these standard pre- knee arthroplasty (38 unicompartmental operative contraindications had no influence on the successful outcome ofthe knee arthroplasties), and 2 patients un- procedure using the Oxford phase 3 device. derwent simultaneous unicompartmental knee arthroplasty with contralateral total knee arthroplasty. Twenty-nine patients ince the inception of unicondylar knee the percentage of patients with osteoarthritic underwent staged procedures (58 unicom- arthroplasty, clinicians have debated knees who may be candidates for unicom- partmental knee arthroplasties) Sthe ideal indications for this conserva- partmental knee arthroplasty has been report- Among the patients were 156 (58%) tive procedure. Despite being reported more ed to be between 2% and \5%.^-^ women, and surgery was performed on than 15 years ago, many surgeons continue In April 2004, the US Food and Drug the right knee in 163 patients of 318 (51%) to follow the standard indications outlined by Administration approved the use of the Ox- knees. The average patient age was 61.6 Kozinn and Scott.' These restrictive indica- ford Phase 3 unicondylar prosthesis (Biomet years (range: 33.6-90.1 years; SD: 10.7). The tions and contraindications exclude patients Orthopedics, Inc, Warsaw, Ind) for implanta- who weigh >82 kg (187 lbs), are younger tion in unicompartmental knee arthroplasty. Drs Berend and Lombardi and Ms Adams are from Joint Implant Surgeons, Inc, and Drs Berend than age 60 years, and who have more than Using more liberal indications, the implant and Lombardi are from the Mount Carmel New Alba- minimal erosive changes in the pateUofemo- has proven to be a conservative yet long-last- ny Surgical Hospital and The Ohio State University, ral articulation.' Anterior knee pain, thought ing intervention for anteromedial arthritis New Albany, Ohio. to be a sign of significant patellofemoral in- of the knee.''^ The purpose of this study is Drs Berend and Lombardi are consultants for the sponsor. They are also stockholders and receive volvement, has also been an exclusion criteri- to investigate the early outcomes of Oxford financial support from them. Ms Adams has no fi- on. Applying these conservative indications. medial unicompartmental knee arthroplasty nancial interests in the materials mentioned herein.

MAY 2007 I Volume 30 • Number 5/SUPPLEMENT 19 Feature Article

Table Outcomes by Preoperative Demographic Variables

Knee Society Knee Society Log-Rank/ Wilcoxon Variable Pain Score Total Score Significance Survival Ratellofemoral disease Present 43 Absent 41 Present 91 Absent 89 P=0.52 P=0.74 Not significant P=0.04 Young age 50 y <50 y, 36 >50 y, 43 <50 y, 84 >50 yrs, 91 P=0.002 P=0.79 Not significant P=0.002 Young age 60 y <60 y, 41 >60 y, 43 <60 y, 89 >60 y, 91 P=0.002 P=0.80 Not significant P=0.08

Weight 82 kg >82, kg 41 <82 kg, 43 >82 kg, 90 <82 kg, 91 P=0.07 P=0.82 Not significant P=0.24

Weight 90 kg >90, kg 41 <90 kg, 43 >90 kg, 89 <90 kg, 91 P=0.06 P=0.30 Not significant P=0.09

BMI 32 kg/m^ >32, 42 <32, 42 >32, 90 <32, 90 P=0.5 P=0.30 Not significant P=0.5

BMI 35 kg/m^ >35, 40 <35, 43 >35, 88 <35, 91 P=0.04 P=0.09 Not significant P=0.03

BMI 40 kg/m2 >40, 36 <40, 43 >40, 83 <40, 91 P=0.002 P=0.30 Not significant P=0.01

Isolated medial knee Not isolated 41 Isolated 43 Not isolated, 89 Isolated 91 P=0.09 P=0.50 Not significant pain P=0.2

Abbreviation: BMI=body mass index

patient was aged <60 years in 47.5% of the Patients were asked preoperatively to Cheshire, United Kingdom). Parametric procedures (n = 151), and the patient was describe the area of their most severe pain as analysis was performed using an unpaired aged <50 years in 14.5% of the procedures medial, lateral, anterior, posterior, or global. Student's t test. Multiple groups were ana- (n = 46). Average height was 1.7 m (5.53 ft) Patients who described more than two areas lyzed using analysis of variance (ANOVA). (range: 1.42-1.85 m [4.67-6.33 ft]; SD: 0.1 were categorized as global. Isolated medial Survivorship curves were constructed using [4.1]). Average weight was 92 kg (202 lbs) pain was reported in 211 (67%) knees, ante- the Kaplan-Meier method, and survivorship (range: 51-158 kg [112-348 lbs]; SD: 18 rior knee pain in 20 (6%) knees, and lateral between groups was compared using log- [39.6]). In 220 knees (69%), the patient's pain in 9 (3%) knees. Posterior knee pain was rank and Wilcoxon methods. Correlations weight was >82 kg. The average body mass reported in 9 (3%) knees and global pain in were examined using Spearman's rank and index (BMI) of patients in the study was 65 (21%) knees. Pain location was not re- Kendall rank tests. AU analyses were per- 32.3 kg/m^ (range: 20-58 kg/m^; SD: 6). In ported for 4 knees. formed using 95% confidence intervals, and 83 cases (26%), the patients were catego- Blinded radiographs of standardized an- a P value of <.O5 was considered significant. rized as obese with BMI >35. Patients were teroposterior standing, true lateral, and sun- morbidly obese with a BMI >40 kg/m^ in rise patellofemoral views were reviewed. The RESULTS 30 of the cases (9%). One hundred fifty presence and severity of radiographic patel- Average follow-up was 8 months (range: (47%) unicompartmental knee arthroplas- lofemoral disease on the lateral and sunrise 6 weeks to 28 months). There were six re- ties were performed in patients with a BMI views were recorded. vision surgeries in this group for an early >32 kg/m^, a category considered at risk Statistical evaluation was performed us- survivorship of 98.1%. One knee was re- for failure secondary to obesity.-^' ing StatsDirect software (Stats Direct, Ltd, vised for ongoing pain, one for tibial plateau

20 ORTHOPEDICS I www.ORTHOSuperSite.com IDENTIFYING UNICONDYLAR ARTHROPLASTY PATIENTS I BEREND ET AL

Age Less Than 50 Years fracture, and three for tibial loosening. One Age Less Than GO Years knee was revised in a two-stage manner for infection. The average American Knee Soci- O Over SO \tm ety (AKS) pain score at early follow-up was • Under SO Yean 42/50, and the overall AKS score averaged 90/100. Range of motion (ROM) averaged 118°. ROM >115° was obtained in 262 knees (82%). When stratified by aged <50 years, there Time (months) was no difference in early failure (Figure 1:

Log-rankAVilcoxon; P=ns). Higher average Figure 1: Kaplan-Meier survival analysis shows no difference in cumulative survival of unicompartmental AKS pain scores were seen in patients aged knee arthroplasty in patients stratified by age younger or older than 50 years. Figure 2: Kaplan-Meier >50 years (43 versus 36; P=.OO2). Average survivai analysis shows no difference in cumulative survival of unicompartmental knee arthroplasty in AKS total scores were also higher in the old- patients stratified by age younger or older than 60 years. er than 50 age group (91 versus 84; P=.OO2). BMI Greater Than 32 BMI Greater Than 35 Age younger than 60 years was also not pre- 1.00' r'-''"'-S^;;:v:^,..: ^ ""^^^^^'H-^":'.".:"....."'""'- '• -'"^— dictive of early failure (Figure 2: Log-Rank/

"S 0.7S- O Leu than 35 Wilcoxon; P=ns). Higher average AKS pain OUulhan32 • Crcater than 33 scores were observed in the patient group a Greater than 32 over age 60 (41 versus 43; P=.O3). The aver-

age AKS total scores were not significantly Cumulativ e Surviva l Cumulativ e S different between the groups of patients older and aged <60 years. These data are shown in 0 to 20 10 20 Time (rrtonths) the Table. Ttme (months) Weight was not predictive of stirvivorship El or clinical outcome when stratified by weight Figure 3; Kaplan-Meier survivai anaiysis shows no difference in cumulative survival of unicompartmentai >82 kg or weight >90 kg. When stratified knee arthroplasty in patients stratified by BMI greater or less than 32 kg/m^. Figure 4; Kaplan-Meier sur- vival analysis shows no difference in cumulative survivai of unicompartmental knee arthroplasty patients by BMI, no differences in early survival or stratified by BMi greater or less than 35 kg/m^. revision were observed. Survivorship in pa- tients with BMI>32 kg/m^ (Figure 3), obese patellofemoral joint arthritis on preoperative difference in AKS clinical scores between patients with a BMI>35 kg/m^ (Figure 4), radiographs (91 versus 89; P=.O4). These the different sites of preoperative pain and morbidly obese patients with a BMI>40 data are shown in the Table. No correlation (P=.34 and P=.46). kg/m^ (Figure 5) was no different between could be established between the presence or the groups (P=ns). Clinical outcomes were severity of pre-existing patellofemoral joint DISCUSSION significantly better in the <35 BMI group disease and the location of preoperative pain In patients with anteromedial disease, in- and the <40 BMI group. The Table shows the described by the patient. dicated by an intact anterior cruciate ligament average AKS pain and total score, and Lx)g- Isolated medial knee pain was not a sig- and fiinctionally normal medial collateral rank/Wilcoxon analysis for survivorship be- nificant predictor of survivorship or failure ligament with a correctable varus deformity, tween these groups. with the device (Figure 7: Log-Rank/Wil- the results of Oxford unicompartmental knee Survivorship in knees with pre-existing coxon; P=ns). American Knee Society pain arthroplasty have been excellent.'-'"'* There radiographic signs of patellofemoral joint scores in knees with isolated medial preop- are certain indications and contraindica- disease was not significantly different than erative pain was not significantly different tions, however, that are widely debated.''^"^^ in knees with no preoperative patellofemoral than AKS pain scores in knees with pain Kozinn and Scott' are credited with describ- radiographic changes (Figure 6). There was other than medially based pain (43 versus 41; ing the so-called standard indications for uni- a trend toward a higher average AKS pain P=.O9). American Knee Society total scores compartmental knee arthroplasty. The Ox- score in knees with pre-existing patellofemo- were also not significantly different between ford indications do not routinely use weight, ral joint disease (43 versus 41; P=.O52). In the isolated medial sided pain and the non- age, activity, the presence of patellofemoral knees with pre-existing patellofemoral joint isolated medial sided pain groups (91 versus disease, or the location of preoperative pain disease, the average AKS total score was 89; P=. 16). These data are shown in the Table. as indications or contraindications for the significantly higher than in knees without One-way ANOVA did not reveal a significant procedure. In the short term, young age.

MAY 2007 |Voiume30 • Number 5/SUPPLEMENT 21 Feature Article

BMI Greater Than 40 Patellofemoral Joint Disease Isolated Medial Sided Knee Pain 1.00 •

OLcwtlunW oDisease oNotbobted D Cfwder than 40 QNoObeasc D bobted Medial Cumulativ e S u g

) 10 20 Time (months) Time (months) 0 Time (months) Figure 5: Kaplan-Meier survival anaiysis shows no difference in cumuiative survival of unicompartmenfai knee arthroplasty in patients stratified by BiVli greater or less than 40 kg/m^. Figure 6: Kapian-Meier survivai anaiysis shows no difference in cumulative survival of unicompartmental knee arthroplasty in knees stratified by presence or absence of pateliofemorai joint disease. Figure 7: Kaplan-Meier survival anaiysis shows no difference in cumuiative survivai of unicompartmentai knee arthropiasty in knees stratified by pain isolated to or more extensive than the mediai compartment. weight or obesity, radiographic patellofemo- also be more pain in the early postoperative patellofemoral joint. Significant involvement ral disease, and the location of preoperative period because forces and demands on the of a second compartment, namely the patel- pain do not appear to significantly affect the prosthesis and operative leg are significantly lofemoral joint, is considered a contraindica- outcomes of the procedure. higher in the obese patient. Despite this, no tion to unicompartmental knee arthroplasty.' Weight >82 kg was originally described significant differences in survivorship or fail- Despite past consideration of patellofemoral as a contraindication to fixed-bearing uni- ure were noted in the obese patients. joint arthritis as a contraindication, the re- compartmental knee arthroplasty.' Reviewed There are few alternatives for the treat- sults of unicompartmental knee arthroplasty individually, weight >82 kg or >90 kg was ment of medial compartmental arthritis in with the Oxford device have been outstand- not associated with failure or decreased clini- young adults. High tibial osteotomy lost ing and progression of disease in the patel- cal outcomes in the current series, despite a favor in recent years because the long-term lofemoral joint has not presented as a clinical large percentage of patients in these catego- results have been unsatisfactory." Although problem.'*"'^'^'''^' This has not been the case ries. In one series examining a resurfacing, no increased risk of early failure of the device with fixed-bearing designs. Recent reports fixed-bearing type unicompartmental knee was seen in our patients younger than age indicate that progression of patellofemoral arthroplasty, BMI >32 kg/m^ was a signifi- 50 or 60 years, slightly lower clinical scores joint disease may be a significant cause of cant predictor of early failure.'^' Although the were observed in these patients. This may be second-decade failure with certain fixed- follow-up in the current series is short- term, attributed to increased demands on the device bearing designs.*'^^'^' In comparison, the previous failures described by Berend et al^' or higher expectations in these younger, more long-term reports of unicompartmental knee occurred early and would have been seen in active patients. Again, with traditional fixed- arthroplasty with the Oxford phase 3 unicon- the follow-up period reported herein. Again, bearing unicompartmental knee arthroplasty, dylar prosthesis do not indicate progression nearly half of the patients in the current series failure in younger patients is predominantly of patellofemoral joint arthritis as a signifi- (47%) had a BMI >32 kg/ml A recent report associated with polyethylene wear. However, cant failure with this device.'^'^'^-^'' We did also supported the contention that weight or good results with some fixed-bearing designs not consider the radiographic and visual ap- BMI is not an independent predictor of sur- have been obtained in younger patients.^-^^ pearance of the patellofemoral joint intraop- vivorship in the mid to long term.^^ There In this population, survivorship has been ex- eratively in using the Oxford phase 3 pros- is concern that in the long term, increasing cellent, with very little wear reported using thesis in unicompartmental knee arthroplasty, weight may lead to increased wear of the the fully congruent Oxford mobUe-bearing and the current report does not indicate any bearing. Our study does not address this issue; design.^'*'^ Unicompartmental knee arthro- negative effect. In addition, we were unable however, weight was not a predictor of in- plasty with the Oxford phase 3 unicondylar to establish a relationship between the pres- creased wear with the Oxford fully congruent prosthesis has achieved 10-year outcomes ence of radiographic patellofemoral joint dis- mobile bearing.^'*'^^ We noted a significantly and survivorship in patients aged <60 years ease and the presence of preop)erative anterior higher AKS pain and total score in the subset that is similar to results with older, more sed- knee pain. In the short term, we have not seen of patients with a BMI lower than 35 kg/m^ entary individuals.^" any patients with moderate or severe anterior or 40 kg/m^. Given the early follow-up, this Perhaps the most controversial con- knee pain postoperatively. In fact, there was likely represents difficulty with rehabilitation traindication for unicompartmental knee a nearly significantly higher AKS pain score and mobility in obese individuals. There may arthroplasty is the preoperative status of the in patients with pre-exisdng patellofemoral

22 ORTHOPEDiCS I www.ORTHOSuperSite.com IDENTIFYING UNICONDYLAR ARTHROPLASTY PATIENTS I BEREND ET AL

joint disease, and the total AKS was signifi- knee arthroplasty: an evaluation of selection cri- independent 10-year survival analysis. Knee teria. Clin Orthop Relat Res. 1993;286:143-148. cantly higher in those patients with pre-exist- Surg Sports Traumatol Arthrosc. 2006; 14:40-45. 3. Laskin RS. Unicompartmental knee replace- ing disease. 19. Rees JL, Price AJ, Lynskey TG, et al. Medial ment: some unanswered questions. Clin Orthop unicompartmental arthroplasty after failed high The location of preoperative knee pain Relat Res. 2001; 392:267-271. tibial osteotomy. J Bone Joint Surg Br. 2001; did not predict successful outcomes using 4. Sculco TP. Orthopedic cross-fire: can we jus- 83:1034-1036. the Oxford phase 3 prosthesis. We did note, tify unicondylar arthroplasty as a temporizing 20. Price AJ, Dodd CA, Svarf UG, Mumy DW. Ox- procedure? In opposition. J Arthroplasty. 2002; ford medial unicompartmental knee arthroplasty however, that the majority of patients with 17(suppl4):56-58. in patients younger and older than 60 years of anteromedial disease did have isolated me- 5. Naudie D, Guerin J, Parker DA, Boume RB, age. 7Boneyo!n/5M/sBr. 2005; 87:1488-1492. dial pain.'^ Pain in other locations was not Rorabeck CH. Medial unicompartmental knee 21. Berend KR, Lombaidi AV Jr, Mallory TH, Ad- associated with a lower AKS pain score or arthroplasty with the Miller-Galante prosthesis. ams JB, Groseth KL. Early failure of minimally J Bone Joint Surg Am. 2004; 86:1931-1935. invasive unicompartmentzJ knee arthroplasty is AKS total score or failure in the current se- 6. Gesell MW, Tria AJ Jr. MIS unicondylar knee associated with obesity. Clin Orthop Relat Res. ries. These data and the data on patellofemo- arthroplasty: surgical approach and early results. 2005; 440:60-66. ral joint disease may be biased by surgeon se- Clin Orthop Relat Res. 2004; 428:53-60. 22. Tabor OB Jr, Tabor OB, Bemaid M, Wan JY. lection. Without a randomized, blinded study, 7. Ritter MA, Faris PM, Thong AE, Davis KE, Unicompartmental knee arthroplasty: long-term success in middle-age and obese patients. J Surg the exact role of patellofemoral joint arthritis Meding JB, Berend ME. Intra-operative findings of carus osteoarthritis of the knee: an analysis of Orthop Adv. 2005; 14:59-63. and anterior knee pain cannot be determined. pre-operadve alignment in potential candidates 23. Pagnano MW, Clarke HD, Jacofsky DJ, et al. The current study, however, does not consider for unicompartmental arthroplasty. J Bone Joint Surgical treatment of the middle-aged patient Surg Br. 2004; 86:43-47. patellofemoral joint disease and anterior knee with arthritic knees. Instr Course Lect. 2005; 54:251-259. pain contraindications in otherwise acceptable 8. Berger RA, Meneghini RM, Sheinkop MB, et al. The progression of patellofemoral arthrosis after 24. Argenson JN, O'Connor JJ. Polyethylene wear candidates for the procedure. medial unicompartmental replacement: results in meniscal knee replacement: a one to nine-year The current study provides only short-term at 10-15 years. Clin Orthop Relat Res. 2004; retrieval analysis ofthe oxford knee. JBone Joint 428:92-99. r. 1992; 74:228-232. data, and longer follow-up is needed to ensure 9. Goodfellow J,O'ConnorJ,Murray DW. The 25. Price AJ, Short A, Kellett C, et al. Ten-year in that the Oxford phase 3 prosthesis is as suc- Oxford meniscal unicompartmental knee. vivo wear measurement of a ftilly congruent mo- cessful in US patients as it has been in patients J Knee Surg. 2002; 15:240-246. bile bearing unicompartmental taiee arthroplasty. abroad. There is no doubt that a rapid and 10. Munray DW, Goodfellow JW, O'Connor JJ. The J Bone Joint Surg Br. 2005; 87:1493-1497. functional retum to activities is possible after Oxford medial unicompartmental arthroplasty: 26. Kebiish PA, Baird JL. Mobile-bearing unicom- a ten-year survival study. J Bone Joint Surg Br. partmental knee arthroplasty: a 2-center study unicompartmental knee arthroplasty in appro- 1998; 80:983-989. with an 11-year follow-up. J Arthroptasty. priately selected patients."'^ Survivorship at 11. Price AJ, Webb J, Topf H, et al. Rapid recovery 2004;19(suppl2):87-94. 7 years is 97.3% with this device.^" The con- after Oxford unicompartmental arthroplasty 27. Rajasekhar C, Das S, Smith A. Unicompart- through a small incision. J Arthroplasty. 2001; servative nature of bone and ligament preser- mental knee arthroplasty: 2-12-year results in a 16:970-976. community hospital. J Bone Joint Surg Br. 2004; vation and long-term survivorship of the Ox- 12. Beard DJ, Murray DW, Rees JL, Price AJ, Dodd 86:983-985. ford phase 3 device warrant its position at the CA. Accelerated recovery for unicompartmental 28. Berger RA, Delia Valle CJ, Jacobs JJ, Sheinkop top of the list of altematives for treating me- knee replacement—a feasibility study. Knee. MB, Rosenberg AG, Galante JO. The progres- 2002; 9:221-224. sion of patellofemoral arthrosis after medial dial disease in the growing 40- to 60-year-old 13. Goodfellow J, O'Connor J. Clinical results with unicompartmental replacement: results at 10- population.^'*-^^'''' In addition, by restoring the the Oxford knee. Clin Orthop Relat Res. 1986; 15 years. Clin Orthop Relat Res. 2006; knee to the predisease state with ligamentous 205:21-42. 452:285-286. balance and fianctional kinematics, activity 14. Goodfellow J, O'Connor J. The anterior cruciate 29. Berger RA, Meneghini RM, Jacobs JJ, et al. Results of unicompanmental knee arthroplas- and lifestyle are considerably improved to ligament in knee arthroplasty: a risk factor with unconstrained meniscal prosthesis. Clin Orthop ty at a minimum of 10-years follow-up. J Bone near normal in many patients following this RelatRes. 1992;276:245-252. Joint Surg Am. 2005; 87:999-1006. conservative procedure.^''^^ The current series 15. White S, Ludkowski PF, Goodfellow J. Antero- 30. Pandit H, Jenkins C, Barker K, Dodd CA, supports the elimination of age, weight, pres- medial osteoarthritis of the knee. / Bone Joint Murray DW. The Oxford medial unicompart- SurgBr. 1991; 73:582-586. mental knee replacement using a minimally- ence of patellofemoral joint disease, and ante- invasive approach. J Bone Joint Surg Br 2006; rior knee pain from the list of contraindica- 16. Price AJ, Waite JC, Svard U. Long-term clinical 88:54-60. results of the medial Oxford unicompartmental tions to unicompartmental knee arthroplasty knee arthroplasty. Clin Orthop Relat Res. 2005; 31. Fisher N, Agarwal M, Reuben SF, Johnson when the Oxford phase 3 device is used. H 435:171-180. DS, Tumer PG. Sporting and physical activity following Oxford medial unicompartmental 17. Svard UC, Price AJ. Oxford medial unicompart- knee arthroplasty. Knee. 2006; 13:296-300. REFERENCES mental knee arthroplasty. A survival analysis of an independent series. J Bone Joint Surg 32. Walton NP, Jahromi I, Lewis PL, Dobson PJ, 1. Kozinn SC, Scott R. Unicondylar knee arthro- Br. 2001; 83:191-194. Angel KR, Campbell DG. Patient-perceived . plasty. J Bone Joint Surg Am. 1989; 71:145-150. outcomes and retum to sport and work: TKA 18. Vorlat P, Putzeys G, Cottenie D, et al. The versus mini-incision unicompartmental knee 2. Stem SH, Becker MW, Insall JN. Unicondylar Oxford unicompartmental knee prosthesis: an arthroplasty. Knee. 2006; 19:112-116.

MAY 2007 I Volume 30 • Number 5/SUPPLEMENT 23 STATE OF THE ART REVIEW BMJ: first published as 10.1136/bmj.n1648 on 26 July 2021. Downloaded from Long covid—mechanisms, risk factors, and ­management Harry Crook,1 Sanara Raza,1 Joseph Nowell,1 Megan Young,1 Paul Edison1,2

ABSTRACT

1 Since its emergence in Wuhan, China, covid-19 has spread and had a profound Faculty of Medicine, Imperial College London, London, UK effect on the lives and health of people around the globe. As of 4 July 2021, more 2 Cardiff University, Cardiff, UK than 183 million confirmed cases of covid-19 had been recorded worldwide, Correspondence to: P Edison [email protected] and 3.97 million deaths. Recent evidence has shown that a range of persistent Cite this as: BMJ 2021;374:n1648 http://dx.doi.org/10.1136/bmj.n1648 symptoms can remain long after the acute SARS-CoV-2 infection, and this condition Series explanation: State of the is now coined long covid by recognized research institutes. Studies have shown Art Reviews are commissioned on the basis of their relevance that long covid can affect the whole spectrum of people with covid-19, from those to academics and specialists with very mild acute disease to the most severe forms. Like acute covid-19, long in the US and internationally. For this reason they are written covid can involve multiple organs and can affect many systems including, but predominantly by US authors. not limited to, the respiratory, cardiovascular, neurological, gastrointestinal, and musculoskeletal systems. The symptoms of long covid include fatigue, dyspnea, cardiac abnormalities, cognitive impairment, sleep disturbances, symptoms of post- traumatic stress disorder, muscle pain, concentration problems, and headache. This http://www.bmj.com/ review summarizes studies of the long term effects of covid-19 in hospitalized and non-hospitalized patients and describes the persistent symptoms they endure. Risk factors for acute covid-19 and long covid and possible therapeutic options are also discussed. on 11 August 2021 at Atlantic Health Hospitals. Protected by copyright. Introduction who have had covid-19, and this condition is known Coronavirus disease 2019 (covid-19) has spread as long covid. The National Institute for Health across the world. As of 4 July 2021, more than 183 and Care Excellence (NICE) defines long covid as million confirmed cases of covid-19 have been the symptoms that continue or develop after acute recorded worldwide, and more than 3.97 million covid-19 infection and which cannot be explained deaths have been reported by the World Health by an alternative diagnosis. This term includes Organization .1 The clinical spectrum of covid-19 ongoing symptomatic covid-19, from four to 12 ranges from asymptomatic infection to fatal disease.2 weeks post-infection, and post-covid-19 syndrome, 3 The virus responsible for causing covid-19, severe beyond 12 weeks post-infection.9 Conversely, acute respiratory syndrome coronavirus 2 (SARS- The National Institutes of Health (NIH) uses the CoV-2), enters cells via the angiotensin-converting US Centers for Disease Control and Prevention enzyme 2 (ACE2) receptor.4 Once internalized, (CDC) definition of long covid, which describes the virus undergoes replication and maturation, the condition as sequelae that extend beyond four provoking an inflammatory response that involves weeks after initial infection.10 People with long covid the activation and infiltration of immune cells by exhibit involvement and impairment in the structure various cytokines in some patients.5 The ACE2 and function of multiple organs.11-14 Numerous receptor is present in numerous cell types throughout symptoms of long covid have been reported and the human body, including in the oral and nasal attributed to various organs, an overview of which mucosa, lungs, heart, gastrointestinal tract, liver, can be seen in fig 1. Long term symptoms following kidneys, spleen, brain, and arterial and venous covid-19 have been observed across the spectrum of endothelial cells, highlighting how SARS-CoV-2 can disease severity. This review examines the long term cause damage to multiple organs.6 7 impact of symptoms reported following covid-19 The impact of covid-19 thus far has been infection and discusses the current epidemiological unparalleled, and long term symptoms could have understanding of long covid, the risk factors that a further devastating effect.8 Recent evidence may predispose a person to develop the condition, shows that a range of symptoms can remain after and the treatment and management guidelines the clearance of the acute infection in many people aimed at treating it. the bmj | BMJ 2021;374:n1648 | doi: 10.1136/bmj.n1648 1 STATE OF THE ART REVIEW BMJ: first published as 10.1136/bmj.n1648 on 26 July 2021. Downloaded from

+HDUW &KHVWSDLQV 0\RFDUGLDO LQƠDPPDWLRQ ↑6HUXPWURSRQLQ 3DOSLWDWLRQV

/XQJV %UDLQ '\VSQHD Š%UDLQIRJš &KHVWSDLQ 'HOLULXP &RXJK )DWLJXH 6OHHSGLVWXUEDQFHV 'HSUHVVLRQDQ[LHW\376'2&6 $&( $&( $&( 3DQFUHDV 6SOHHQ

3DQFUHDWLFLQMXU\ $&( ↓ 7 %O\PSKRF\WH 3DQFUHDWLWLV $WURSK\RIO\PSKRLGIROOLFOHV

$&( $&( /LYHU

$&(  /LYHULQMXU\

$&(  ↑$VSDUWDWH .LGQH\ $&( DPLQRWUDQVIHUDVH 5HQDOLPSDLUPHQW ↑$ODQLQHDPLQRWUDQVIHUDVH $FXWHNLGQH\LQMXU\ http://www.bmj.com/

*DVWURLQWHVWLQDOWUDFW %ORRGYHVVHOV 'LDUUKHD ,QƠDPPDWLRQ 1DXVHD 9HVVHOGDPDJH 6RUHWKURDW &RDJXORSDWK\ 0LFURDQJLRSDWK\ on 11 August 2021 at Atlantic Health Hospitals. Protected by copyright.

Fig 1 | Multi-organ complications of covid-19 and long covid. The SARS-CoV-2 virus gains entry into the cells of multiple organs via the ACE2 receptor. Once these cells have been invaded, the virus can cause a multitude of damage ultimately leading to numerous persistent symptoms, some of which are outlined here

Methods exclusion of studies, limited exclusion and inclusion We searched PubMed and Embase databases for criteria were applied. We excluded papers that were articles published between January 2020 and not relevant to or did not mention long covid, while May 2021. Our search terms were “long covid” or studies mentioning long covid in any capacity were “post-covid-19” or “COVID long-haulers” or “SARS- initially included owing to the novelty of the field. CoV-2” and “epidemiology” or “fatigue” or “fatigue Furthermore, we considered long covid studies syndrome” or “dyspnoea” or “breathlessness” or regardless of their cohort sizes or study design. We “shortness of breath” or “cardiac” or “cardiovascular” discovered and read fully 227 articles on long covid, or “heart” or “cognition” or “cognitive impairment” and we discussed each to determine which would or “mental health” or “depression” or “anxiety” be included in the finalized article. We performed or “psychiatric” or “central nervous system” or further manual searching for additional articles “autonomic nervous system” or “isolation” or and treatment guidelines using relevant databases, “loneliness” or “sleeplessness” or “sleep” or “smell” including nice.org.uk and clinicaltrials.gov. In total, or “taste” or “olfactory” or “gustatory” or “risk 218 references were included. Studies examining factors” or “treatment”. To avoid unintentionally long covid are limited, therefore limited exclusion removing articles, no filters were applied. We criteria were applied. retrieved 61 881 articles in the first instance. To screen articles, titles were read by authors first, Studies of long covid followed by abstracts to further narrow down the Studies have assessed people who have had covid-19 number of records considered. To avoid unnecessary to examine the symptoms associated with long covid.

2 doi: 10.1136/bmj.n1648 | BMJ 2021;374:n1648 | the bmj STATE OF THE ART REVIEW BMJ: first published as 10.1136/bmj.n1648 on 26 July 2021. Downloaded from

These studies are summarized in table 1. The articles many patients likely to develop long covid and require included throughout this review were selected in long term support and treatment. Further studies favor of quality, with large observational studies of are required to consolidate our epidemiological greatest interest. Most of the studies included are understanding of long covid. cross-sectional or cohort observational studies with large cohorts; however, because of the novelty of Covid-19 variants of concern the disease and paucity of data, studies involving Since the start of the pandemic, several covid-19 smaller cohorts and case series were also included. variants have emerged that have an increased Any patient with covid-19 may develop long covid, transmissibility and may result in more severe acute regardless of the severity of their infection and the disease. In the UK, one of the first variants of concern intensity of the treatment they received. Patients to appear was the so called “Kent variant,” from treated on wards and intensive care units (ICUs) show the B.1.1.7 lineage, now termed the Alpha variant. little difference in incidence of long term symptoms This variant has approximately 50% increased associated with covid-19.17 The proportion of people transmissibility54 and likely increases acute disease that develop long covid symptoms, whether they are severity.55 As of 30 June 2021, the Alpha variant treated with oxygen alone, with continuous positive has been confirmed in more than 275 000 cases airway pressure, or with invasive ventilation, is in the UK56 and spread to at least 136 countries similar.16 Many patients with mild acute symptoms around the world.57 Other variants of concern or also develop long covid symptoms,13 in fact, studies under investigation include the Beta, Gamma, Zeta, show minimal differences between the prevalence of Theta, and Kappa variants.56 The CDC reports the long covid symptoms between hospitalized and non- emergence of variants of concern and interest in the hospitalized covid-19 patients.19 US.58 New covid-19 variants will continue to emerge and spread as we progress through the pandemic, for Epidemiology example, the Eta and Delta variants have arisen, with The reported incidence and mortality rates of over 161 000 cases of the rapidly spreading Delta covid-19 vary between countries, making it difficult variant confirmed in the UK, as of 30 June 2021.56 to accurately predict the number of patients who Recently, the Lambda variant has emerged, which will progress to long covid. Similarly, the accurate will require close monitoring. The ability of these http://www.bmj.com/ reporting of long covid is complicated. The disparity viral strains to inflict long term complications needs in this epidemiological data is likely the result of to be examined fully. To speculate, it may be that one several factors, including differences in the base variant causes more damaging long term effects than population, the accuracy of diagnosis, the reporting others and, therefore, patients infected with such a systems, and the capability of healthcare systems. variant who go on to develop long covid symptoms Although determining the exact epidemiological data may require additional support, as well as more rapid

of long covid is difficult, this information is needed and intense treatment strategies to combat their long on 11 August 2021 at Atlantic Health Hospitals. Protected by copyright. to inform healthcare systems and governments when term symptoms. developing support and treatment algorithms. The volume of published literature describing cases of Long covid definition patients with covid-19 who subsequently develop Long covid gained widespread attention following long covid symptoms is continually growing, which an account published on 5 May 2020 in BMJOpinion will allow for an improved understanding of its where an infectious disease professor shared his epidemiology. experience of seven weeks on a “rollercoaster of ill The current disparities between long covid health” following covid-19.59 The patient-made term epidemiology reporting are owing to many reasons, long covid was then made popular following the rise including the length of follow-up period, population in the use of #LongCovid on Twitter.60 This, plus the assessed, accuracy of self-reporting, and symptoms growing number of peer reviewed articles published examined. Studies around the world have reported since, has highlighted a post-covid-19 syndrome that various incidence rates for long covid with different can last for many weeks after the acute infection. follow-up examination times after the acute Long covid is now a recognized term in scientific infection, including 76% of people at 6 months,50 literature. The NICE guidelines on managing the long 32.6% at 60 days,51 87% at 60 days,15 and 96% at term effects of covid-199 and the CDC10 define long 90 days.52 These finding are not fully corroborative, covid patients or covid long haulers as individuals but they show that a substantial proportion of people with ongoing symptoms of covid-19 that persist who have had covid-19 may develop long covid. The beyond four weeks from initial infection. UK Office for National Statistics (ONS) has released 53 data on the prevalence of long covid symptoms. Symptoms They estimated that the five week prevalence of Fatigue any symptom among survey respondents who Fatigue is more profound than being overtired; it tested positive for covid-19 between 22 April and is unrelenting exhaustion and a constant state of 14 December 2020 was 22.1%, while the 12 week weariness that reduces a person’s energy, motivation, prevalence was 9.9%. These figures are worrying for and concentration. Following the SARS outbreak, patients, service providers, and governments, with up to 60% of patients reported ongoing fatigue the bmj | BMJ 2021;374:n1648 | doi: 10.1136/bmj.n1648 3 STATE OF THE ART REVIEW BMJ: first published as 10.1136/bmj.n1648 on 26 July 2021. Downloaded from

Table 1 | Summary of studies that have explored the persisting symptoms post-covid-19 infection, or during long covid Number of Time to subjects in Hospitalized / assessment Study reference study non-hospitalized Study design (average) Symptoms (% of patients) Carfi A, et al, 143 Hospitalized Case series 60.3 days after Fatigue (53.1%); dyspnea (43.4%); joint pain (27.3%); chest pain 202015 onset (21.7%) Mandal S, et al, 384 Hospitalized Cross sectional 54 days post- Fatigue (46.6%); cough (28.6%); breathlessness (56.25%); poor 202016 (analytic) hospital discharge sleep quality (57%) Halpin SJ, et al, 100 Hospitalized (32 ICU treated, Cross sectional 48 days after Fatigue (64%); breathlessness neuropsychological (30%); speech 202017 68 ward treated) (analytic) onset (48%); and swallow (8%) Dennis A, et al, 201 Hospitalized: n=37; Cross sectional 140 days after Fatigue (98%); muscle ache fever (75.1%); chest pain (73.6%); 202013 non-hospitalized: n=164 (analytic) onset (87.6%); shortness of breath sore throat (71.1%); diarrhea (87.1%); headache (82.6%); joint (59.2%) pain (78.1%); Tenforde MW, et 274 Non-hospitalized Cross sectional 14-21 days after Fatigue (38%); cough (46%); congestion (32%); dyspnea (31%); al, 202018 (survey) onset headache (18%); body ache (20%); nausea (13%); sore throat (18%); loss of taste (28%); loss of smell chest pain (20%); abdominal pain (27%); diarrhea (14%); (18%); confusion (20%) Goertz YMJ, et al, 2113 Hospitalized: n=112; non- Cross sectional 79 days after Fatigue (87%);dyspnea (71%); chest tightness (44%);cough (29%) 202019 hospitalized: n=2001 (survey) onset Townsend L, 128 Hospitalized: n=71; non- Cross sectional 72 days after Fatigue (52.3%) et al, 202020 hospitalized: n=57 (analytic) initial symptoms Boscolo-Rizzo P, 187 Non-hospitalized Cross sectional 28 days after Loss of taste or smell (10.6%) et al, 202021 (survey) onset Paderno A, et al, 151 Non-hospitalized Cohort study 30 days after Olfactory dysfunction (17%); gustatory dysfunction (11%) 202022 onset Puntmann VO, 100 Hospitalized: n=33; non- Cohort study 71 days after Cardiac involvement (78%); inflammation (60%); shortness of et al, 202023 hospitalized: n=67 onset troponin levels (71%); breath (36%) ongoing myocardial Helms J, et al, 58 Hospitalized Case series At discharge from Agitation (69%); corticospinal tract delirium development (65%); 202024 hospital syndrome (67%); dysexecutive syndrome (36%) Vaes AW, et al, 1837 Non-hospitalized Cross sectional 79 days after Requirement of personal care (52.4%) 202025 (survey) onset http://www.bmj.com/ Arnold DT, et al, 110 Hospitalized Cross sectional 8-12 weeks after Breathlessness (39%); fatigue insomnia (24%) 202026 (analytic) onset (39%); Cruz RF, et al, 119 Hospitalized Cohort study 4-6 weeks post- Fatigue (67.8%); breathlessness insomnia (56.5%); pain (49.5%) 202027 discharge (32.2%); persistent cough (42.6%); Daher A, et al, 33 Hospitalized Cohort study 6 weeks post- Fever (3%); cough (33%); dyspnea headache (15%); loss of smell 202028 discharge (33%); fatigue (45%); tiredness (12%); loss of taste (9%); diarrhea (45%); sore throat (9%); (9%); angina pectoris (18%) Huang L, et al, 26 Hospitalized Cross sectional Not reported Abnormal cardiac findings (58%); myocardial edema (54%) 202029 (analytic) on 11 August 2021 at Atlantic Health Hospitals. Protected by copyright. Huang Y, et al, 57 Hospitalized Cross sectional At least 30 days Slight cough (10.5%); shortness of occasional wheezing (5.3%) 202030 (analytic) since acute breath (7%); infection Raman B, et al, 58 Hospitalized Cohort study 2-3 months after Lung parenchymal abnormalities fatigue (30/55 55%); 202011 onset (32/53 60.4%); breathlessness Liver injury (11%); renal (36/53 64%); impairment (3%) Savarraj JPJ, et al, 48 Hospitalized Cohort study 90 days after Any neurological symptom (71%); cognitive deficit (BNST) (12%); 202031 onset fatigue (42%); post-traumatic depression symptoms (PHQ-9) stress (PC-PTSD-5 (29%); (11%); anxiety (GAD-7) (9%); Pain sleepiness (17%); (PEG) (64%) Sonnweber T, 109 Hospitalized: n=87; Cohort study 60 days after Iron deficiency (30%); anemia hyperferritinemia (38%) et al, 202032 non-hospitalized: n=22 onset (9.2%); Valiente-De S, et 82 Non-hospitalized Observational 12 weeks after Dyspnea (55.6%); asthenia chills (3.7%); arthromyalgia (2.8%); al, 202033 study onset (44.9%); cough (25.9%); chest hair loss (2.8%); diarrhea (1.9%); pain (25.9%); palpitations (22.2%); anxiety (6.4%); insomnia (1.9%); headache (9.3%); anosmia (9.3%); loss of memory (1.9%); difficulty dysgeusia (5.6%); fever (3.7%) concentrating (1.9%) Sudre CH, et al, 4182 13.9% required hospital Cohort study 28 days after Fatigue (97.7%); headache 202034 treatment, 86.1% required onset (91.2%) no hospital treatment Vaira LA, et al, 138 Hospitalized: n=32; Cohort study 60 days after Smell or taste dysfunction (7.2%) 202035 non-hospitalized: n=106 onset Tomasoni D, et al, 105 Hospitalized Cross sectional 90 days after Smell or taste dysfunction (5.7%); dyspnea (6.7%); fatigue (31.4%); 202036 (analytic) onset gastrointestinal symptoms (1%); cognitive deficits (17.1%) burning pain (10.5%); Mazza MG, et al, 402 Hospitalized: n=300; Cross sectional 30 days post- PTSD (28%); depression (31%); obsessive-compulsive symptoms 202037 non-hospitalized: n=102 (analytic) discharge anxiety (42%); (20%); insomnia (40%) Klein H, et al, 112 Hospitalized: n=6; Cross sectional 6 months after Fatigue (20.5%); smell change memory disorders (5.4%); muscle 202038 non-hospitalized: n=106 (survey) onset (13.4%); breathing difficulty aches (7.14%); headaches (8.9%); taste change (7.1%); (3.57%); hair loss (2.68%) (Continued)

4 doi: 10.1136/bmj.n1648 | BMJ 2021;374:n1648 | the bmj STATE OF THE ART REVIEW BMJ: first published as 10.1136/bmj.n1648 on 26 July 2021. Downloaded from

Table 1 | Continued Number of Time to subjects in Hospitalized / assessment Study reference study non-hospitalized Study design (average) Symptoms (% of patients) Fjaeldstad AW, 204 Non-hospitalized Cross sectional 24 days after Olfactory loss (28/100 28%); gustatory loss (21/104 20%) et al, 202039 (survey) onset Eiros R, et al, 139 Hospitalized: n=23; Cross sectional 10.4 weeks after No symptoms (34%); fatigue piloerection (1%); shortness of 202040 non-hospitalized: n=116 (analytic) onset (27%); anosmia (9%); ageusia breath (26%); chest pain (19%); (5%); headache (5%); sore throat pericarditis-like chest pain (13%); (5%); abdominal pain (4%); palpitations (14%); dizziness (6%); memory loss (3%); joint pain (2%); at least one cardiac symptom (42%) Xiong Q, et al 538 Hospitalized Cohort study 97 days post- General symptoms (49.6%); Cardiovascular related symptoms 202041 discharge physical decline/fatigue (28.3%); (13%); increase in resting heart sweating (23.6%); myalgia (4.5%); rate (11.2%); discontinuous arthralgia (7.6%); chills (4.6%); flushing (4.8%); newly diagnosed limb edema (2.6%); dizziness hypertension (1.3%); psychosocial (2.6%); respiratory symptoms symptoms (22.7%); somnipathy (39%); post-activity polypnea (17.7%); depression (4.3%); (21.4%); non-motor polypnea anxiety (6.5%); dysphoria (1.7%); (4.7%); chest distress (14.1%); feelings of inferiority (0.6%); chest pain (12.3%); sputum (3%); alopecia (28.6%) throat pain (3.2%); Weerahandi H, 152 Hospitalized Cohort study 37 days post- Shortness of breath (74%) et al, 202042 discharge Kamal M, et al, 287 Hospitalized: n=14; Cross sectional Unclear Fatigue (72.8%); anxiety (38%); intermittent fever (11.1%); 202043 non-hospitalized: n=273 (survey) joint pain (31.4%); continuous obsessive-compulsive disorder headache (28.9%); chest pain (4.9%); pulmonary fibrosis (4.9%); (28.9%); dementia (28.6%); diabetes mellitus (4.2%); migraine depression (28.6%); dyspnea (2.8%); stroke (2.8%); renal (28.2%) blurred vision (17.1%); failure (1.4%); myocarditis (1.4%); tinnitus (16.7%); arrhythmia (0.3%) Poyraz BC, et al. 284 Hospitalized: n=112; Cross sectional 50 days following Fatigue (40%); muscle aches light-headedness (7%); numbness http://www.bmj.com/ 202044 non-hospitalized: n=169 (survey) diagnosis (22%); alteration of taste (18%); and tingling sensations on the skin headache (17%); alteration (6%); dyspnea (4%); chest pain of smell (17%); difficulty in (3%); cough (2%); concentration (15%); daytime sleepiness (10%) Landi F, et al, 131 Hospitalized Cohort study 55.8 days after Cough (16.7%); fatigue (51.1%); red eyes (16%); joint pain (25.1%); 202045 onset diarrhea (3.8%); headache shortness of breath (44.2%); loss (10.6%); smell disorder (13.7%); of appetite (9.9%); sore throat dysgeusia (11.4%); (6.8%); rhinitis (14.5%) on 11 August 2021 at Atlantic Health Hospitals. Protected by copyright. Carvalho- 150 Hospitalized: n=53; Cohort study 30 days after Fever (3.6%); shortness of breath weight loss (15.9%); anosmia/ Schneider C, et al, non-hospitalized: n=97 onset (10.7%); chest pain (18%); flu- ageusia (27.8%); heart palpitations 202046 like symptoms (36%); digestive (6.5%); disorders (17.3%); arthralgia (9.8%); cutaneous signs (15.4%) Otte MS, et al, 91 Non-hospitalized Cross sectional 56.55 days after Olfactory impairment (45.1%) 202047 (survey) onset Zhao YM, et al, 55 Hospitalized Cohort study 3 months after Gastrointestinal symptoms exertional dyspnea (14.6%); cough 202048 onset (30.9%); headache (18.2%); and sputum (1.8%) fatigue (16.4%); Frontera JA, et al, 382 Hospitalized Cohort study 6 months post- Fatigue (36%); anxiety (46%); sleep problems (38%); depression 202149 discharge cognitive impairment (50%); (25%); limited activities of daily living (56%)

at 12 months following recovery from the acute fatigue.20 These findings show that fatigue is a major illness.61 In long covid, fatigue is one of the most manifestation of long covid. reported manifestations, with the ONS estimating the five week prevalence of fatigue to be 11.9% Possible mechanisms among people who have had covid-19.53 Fatigue is Chronic fatigue following viral infection may be the a common persisting symptom regardless of severity result of miscommunication in the inflammatory of the acute stage of covid-19. One cross-sectional response pathways62; however, a cross-sectional study found that 92.9% and 93.5% of hospitalized analytical study found no association between and non-hospitalized covid-19 patients, respectively, pro-inflammatory markers and long term fatigue reported ongoing fatigue at 79 days following onset in covid-19 patients with persisting fatigue.20 It of illness.19 Many other cross-sectional and cohort is likely that a range of central, peripheral, and studies report that chronic fatigue is the most psychological factors play a role in the development frequently reported symptom following recovery of post-covid-19 fatigue. A narrative review explains from acute covid-19,15 17 20 27 43 with one showing no that congestion of the glymphatic system and the association between covid-19 severity and long term subsequent toxic build-up within the central nervous the bmj | BMJ 2021;374:n1648 | doi: 10.1136/bmj.n1648 5 STATE OF THE ART REVIEW BMJ: first published as 10.1136/bmj.n1648 on 26 July 2021. Downloaded from

system (CNS), caused by an increased resistance to have no signs of permanent or longlasting lung cerebrospinal fluid drainage through the cribriform damage.28 77 It is likely that only those at high risk plate as a result of olfactory neuron damage, may of developing breathing difficulties, including older contribute to post-covid-19 fatigue.63 people, those who endure acute respiratory distress Hypometabolism in the frontal lobe and cerebellum syndrome, those who have extended hospital stays, has also been implicated in covid-19 patients with and those with pre-existing lung abnormalities, fatigue and is likely caused by systemic inflammation are prone to develop fibrotic-like changes to lung and cell mediated immune mechanisms, rather tissue.78 The fibrotic state observed in some patients than direct viral neuro-invasion.64 65 It is unknown with ongoing dyspnea may be provoked by cytokines whether this finding continues into long covid. such as interleukin-6, which is raised in covid-1979 Negative psychological and social factors and is involved in the formation of pulmonary associated with the covid-19 pandemic have also fibrosis.80 Pulmonary vascular thromboembolisms been linked to chronic fatigue.66 67 Lastly, peripheral have been observed in patients with covid-1981 and factors such as direct SARS-CoV-2 infection of may have detrimental consequences in patients with skeletal muscle, resulting in damage, weakness, and long covid. An overview of the possible mechanism inflammation to muscle fibers and neuromuscular causing dyspnea is outlined in fig 2. junctions may contribute to fatigue.68-71 Overall, it is probable that several factors and mechanisms play Cardiovascular abnormalities a role in the development of post-covid-19 fatigue. Cardiac injury and elevated cardiac troponin levels Figure 2 further outlines these possible mechanisms. are associated with a significantly increased risk of Post-COVID-19 fatigue has been compared with mortality in patients admitted to hospital with acute myalgic encephalomyelitis/chronic fatigue syndrome covid-19 infection.82 83 Persisting cardiovascular (ME/CFS), with many overlaps between the two.72 abnormalities may be burdensome for people Symptoms common to both ME/CFS and long covid with long covid. A cohort study showed cardiac include fatigue, neurological/pain, neurocognitive/ involvement, ongoing myocardial inflammation, and psychiatric, neuroendocrine, autonomic, and elevated serum troponin levels in many people with immune symptoms, with both ME/CFS and long covid-19 at 71 days following diagnosis,23 while a covid patients having long symptom durations, large case series showed that chest pain, possibly http://www.bmj.com/ reduced daily activity, and post-exertional malaise.72 owing to myocarditis, was a common manifestation ME/CFS remains enigmatic, therefore, research into in patients 60.3 days following onset of covid-19 long covid may assist in developing understanding symptoms, with 21.7% of the 143 patient assessed of ME/CFS and vice versa. reporting chest pain.15 Those considered at low risk of severe covid-19, such as young, competitive athletes, Dyspnea have also been found to have residual myocarditis 84 Breathlessness is common in people with long covid. long after recovery from covid-19. In addition to on 11 August 2021 at Atlantic Health Hospitals. Protected by copyright. The ONS estimates that shortness of breath has cardiac complaints, studies have highlighted an a prevalence of 4.6% at five weeks post-covid-19 emerging trend in the development of new onset infection, regardless of presence of acute respiratory postural orthostatic tachycardia syndrome (POTS) symptoms or disease severity.53 Abnormalities in in individuals post-covid-19 infection, because of diffusion capacity for carbon monoxide, total lung autonomic dysfunction.85-89 capacity, forced expiratory volume in the first second, forced vital capacity, and small airway function, have Possible mechanisms been seen in hospitalized covid-19 patients at time of ACE2 receptors are highly expressed in the heart,90 discharge, approximately one month following onset providing a direct route of infection for SARS-CoV-2. of symptoms, showing that lung function in people Studies have shown that sarcomere disruption who have had covid-19 may take time to recover.73 and fragmentation, enucleation, transcriptional Several studies have found that dyspnea is a common changes, and an intense local immune response manifestation following covid-19 infection,16 17 occurs in cardiomyocytes infected by SARS- and one study reported that 43.4% of 143 patients CoV-2.91 92 Pathological responses to acute cardiac assessed were still experiencing dyspnea at 60 days injury and viral myocarditis, such as endothelial after covid-19 onset.15 damage and microthrombosis, can lead to the development of coagulopathy,93 while chronic Possible mechanisms hypoxia and an increase in pulmonary arterial As covid-19 is principally a respiratory illness, acute pressure and ventricular strain may further illness can cause substantial damage to the lungs precipitate the incidence of cardiac injury in people and respiratory tract via SARS-CoV-2 replication who have had covid-19.94 Furthermore, sustained inside endothelial cells, resulting in endothelial immune activation can lead to fibrotic changes95 damage and an intense immune and inflammatory and displacement of desmosomal proteins,96 reaction.74 75 Those who overcome the acute infection which could be arrhythmogenic. Viral infection may develop long term lung abnormalities, leading to has previously been shown to precede POTS97 and, dyspnea76; however, most individuals who develop with the ACE2 receptor expressed on neurons, viral long term breathing difficulties post-covid-19 infection by SARS-CoV-2 may have direct negative

6 doi: 10.1136/bmj.n1648 | BMJ 2021;374:n1648 | the bmj STATE OF THE ART REVIEW BMJ: first published as 10.1136/bmj.n1648 on 26 July 2021. Downloaded from

1

B C Fibroblasts

A

T-cell Alveolus Monocyte ROS Alveolar Cytokines macrophage IL-6, IL-1β TNF-α, IL-8, ROS Blood clot

Neurotoxic factors Basement Collagen and IL-1, TNF-α, ROS membrane damage bronectin

2

B A IL-6, IL-1, TNF-α Autonomic Cardiomyocyte dysfunction death and myositis

C T-cells

Monocytes

D Displacement of desmosomal proteins

Live cardiomyocyte Fig 2 | Long term sequalae of covid-19 (1) In the alveoli Fibrosis of the lungs: (A) Chronic inflammation results in the sustained production of pro-inflammatory cytokines and Dead cardiomyocyte

Increased Pro-in„ammatory reactive oxygen species (ROS) which are released into http://www.bmj.com/ cardiac cytokines bromyoblasts Fibroblast/myobroblast the surrounding tissue and bloodstream. (B) Endothelial damage triggers the activation of fibroblasts, which

3 deposit collagen and fibronectin resulting in fibrotic changes. (C) Endothelial injury, complement activation, A Resting B microglia platelet activation, and platelet-leukocyte interactions, release of pro-inflammatory cytokines, disruption of In„ammatory trigger from Blood normal coagulant pathways, and hypoxia may result covid-19 clot

in the development of a prolonged hyperinflammatory on 11 August 2021 at Atlantic Health Hospitals. Protected by copyright. Self perpetuating neurotoxicity and hypercoagulable state, increasing the risk of thrombosis. (2) In the heart: (A) chronic inflammation Anti in„ammatory cytokines of cardiomyocytes can result in myositis and cause (IL-13, IL-14) E Cognitive impairment Initial anti- cardiomyocytes death. (B) Dysfunction of the afferent in„ammatory C phenotype of Self perpetuating autonomic nervous system can cause complications microglial neurotoxicity Leukocyte activation inltration such as postural orthostatic tachycardia syndrome. (C) Pro-in„ammatory cytokines Prolonged inflammation and cellular damage prompts Excessive/ (TNF-α, IFN-1β, D persistent iNOS) Extravasation of fibroblasts to secrete extracellular matrix molecules and in„ammatory blood derived trigger substances Autonomic collagen, resulting in fibrosis. (D) Fibrotic changes are (???) dysfunction accompanied by an increase in cardiac fibromyoblasts, Activation of pro-in„ammatory microglial while damage to desmosomal proteins results in reduced cell-to-cell adhesion. (3) In the central nervous 4 system: (A) The long term immune response activates glial cells which chronically damage neurons. (B) Psychological Nerve terminal and social factors Hyperinflammatory and hypercoagulable states lead to an increased risk of thrombotic events. (C) Blood-brain Pro-in„ammatory cytokines Fatigue barrier damage and dysregulation results in pathological ROS permeability, allowing blood derived substances and

Muscle ber Neuroin„ammation leukocytes to infiltrate the brain parenchyma. (D) Chronic inflammation in the brainstem may cause autonomic Epimysium dysfunction. (E) The effects of long covid in the brain can Striations Sarcoplasm Perimysium lead to cognitive impairment.(4) Possible mechanisms Muscle ber Z line Z line Z line M line M line causing post-covid-19 fatigue. A range of central, peripheral, and psychological factors may cause chronic fatigue in long covid. Chronic inflammation in the brain,

Myobril as well as at the neuromuscular junctions, may result

Blood vessels Filament A Sarcomere in long term fatigue. In skeletal muscle, sarcolemma band (from Z line to Z line) damage and fiber atrophy and damage may play a role in Atrophy of muscle bers Sarcolemma damage Microthrombi fatigue, as might a number of psychological and social factors the bmj | BMJ 2021;374:n1648 | doi: 10.1136/bmj.n1648 7 STATE OF THE ART REVIEW BMJ: first published as 10.1136/bmj.n1648 on 26 July 2021. Downloaded from

consequences on the autonomic nervous system.98 increased incidences of stroke following covid-19 A complex combination of infection, an autonomic infection. Covid-19 has also been associated with an nervous system induced pro-inflammatoryincreased risk of developing neurological conditions response, and a level of autoimmunity may all including Guillain-Barré syndrome,109 and contribute to the establishment of autonomic neurodegenerative conditions such as Alzheimer’s dysfunction and POTS.89 Figure 2 depicts these disease.110 mechanisms. The pandemic has had a negative effect on mental health, with people who have had covid-19 Cognition and mental health exhibiting long term psychiatric symptoms including Studies have explored cognitive function and post-traumatic stress disorder (PTSD), depression, deficits in patients with covid-19 and suggest that anxiety, and obsessive-compulsive symptoms the virus can cause septic encephalopathy, non- following recovery from the acute infection.36 37 111 112 immunological effects such as hypotension, hypoxia, Quarantine, isolation, and social distancing also have and vascular thrombosis, and immunological damaging effects on mental health and cognition. A effects such as adaptive autoimmunity, microglial rapid review article states that the longer a person is activation, and a maladaptive cytokine profile.99 confined to quarantine, the poorer the outcomes for Additionally, patients admitted to hospital with their mental health,67 while periods of isolation and covid-19 have presented with a range of complaints the inability to work can cause anxiety, loneliness, including encephalopathy, cognitive impairment, and financial concerns, and living through a global cerebrovascular events/disease, seizures, hypoxic health crisis can lead to avoidance behaviors and brain injuries, corticospinal tract signs, dysexecutive behavioral changes.113 The mental health of the older syndrome, an altered mental status, and psychiatric population is greatly affected by social distancing conditions.24 100 101 These findings reveal that and similar measures. By assessing the associations neurological symptoms associated with covid-19 between loneliness, physical activity, and mental are common, diverse, and could pose substantial health both before and during the pandemic, one problems for rehabilitation and ongoing care study found that negative changes of these factors following recovery from covid-19. It is unknown who were not solely owing to longitudinal situations is most affected by cognitive complaints induced by before 2020, therefore the pandemic exerted extra http://www.bmj.com/ covid-19 and how long they persist; however, patient unfavorable effects on loneliness, physical activity, experiences and published summaries of long covid and mental health.114 People living in care homes, have described “brain fog” to be a common and including people with dementia, are vulnerable to debilitating symptom.102-104 covid-19 and to other impacts of the pandemic. Those Critical illness, severe acute respiratory syndrome, with dementia in care homes have been observed and long term ventilator support are known to have to become more depressed, anxious, agitated, and 115 detrimental effects on long term cognition. Before lonely. Protracted social isolation has resulted on 11 August 2021 at Atlantic Health Hospitals. Protected by copyright. the covid-19 pandemic, a retrospective study of 1040 in exacerbation of neuropsychiatric and behavioral ICU treated patients who had respiratory failure, disturbances, including apathy, anxiety, agitation, shock, or both during hospital stays, found that boredom, and confusion in dementia patients living 71% had delirium which lasted around four months in care homes, to a greater degree than for care home following discharge.105 A similar study found that, residents without dementia.116 117 at 3 months post-discharge, 40% of ICU treated Sleeplessness is also commonly reported patients had cognition scores like those of patients following recovery from covid-19, with many studies with moderate traumatic brain injury, while 26% finding poor sleep quality and sleep disturbances had scores similar to patients with mild Alzheimer’s to be frequent following recovery from acute disease. Delirium was also widely reported, with a illness.16 25 31 44 118 119 Furthermore, a retrospective longer duration of delirium associated with worse study of medical records of covid-19 patients treated cognition.106 With many covid-19 patients requiring in Seoul, South Korea, found that after prescriptions ICU admission and mechanical ventilation, long to treat fever, cough, and rhinorrhea, medications term cognitive impairment and delirium are likely to for sleep problems were the next most prescribed pose considerable problems. treatments.120 Knowledge of the covid-19 death Stroke and headache are prevalent in those toll also has a negative impact on quality of sleep, recovered from acute covid-19, with the ONS stress, anxiety, and other negative emotions,121 and estimating the 5 week prevalence of headache sleep problems have been shown to be associated at 10.1% of all covid-19 survivors.13 18 34 43 53 with covid-19 related loneliness.122 This leads us to Exaggerated levels of systemic inflammation, question whether post-covid-19 sleep disturbances observed in some patients as a “cytokine storm,” in are a result of covid-19 infection, the negative effects addition to activation glial cells, poses a substantial of the pandemic, or a combination of both. risk to the brain and increases the likelihood of neurological manifestations including encephalitis Possible mechanisms and stroke.74 Hypercoagulability107 and cardio- Coronaviruses including SARS-CoV-2 can infect the embolisms, formed because of virus related cardiac central nervous system (CNS) via hematogenous or injury,108 are manifestations that could result in neuronal retrograde neuro-invasive routes.123 The

8 doi: 10.1136/bmj.n1648 | BMJ 2021;374:n1648 | the bmj STATE OF THE ART REVIEW BMJ: first published as 10.1136/bmj.n1648 on 26 July 2021. Downloaded from

entry mechanism and subsequent CNS infection may have recovered from covid-19.134-136 Although the explain the high incidence of neuro-inflammation long term effects of covid-19 on the kidneys are not seen in patients with covid-19, and may result in fully elucidated, a study assessing kidney function in damaging long term effects, with associations of viral patients with covid-19 found that 35% had decreased infections and chronic neuro-inflammation with kidney function at 6 months post-discharge.50 neurodegenerative and psychiatric disorders already Acutely, pancreatitis triggered by SARS-CoV-2 has elucidated.123 124 SARS-CoV-2 may also affect the been seen in people with covid-19,6 137 while serum permeability of the blood-brain barrier, which would amylase and lipase levels have been observed to be enable peripheral cytokines and other blood derived higher in people with severe illness compared with substances to enter the CNS and further drive neuro- mild cases, and computed tomography images have inflammation.125 Thrombo-inflammatory pathways shown pancreatic injury.138 A cross sectional study may be the cause of the increased prevalence of stroke found that 40% of patients with covid-19 who were at in covid-19,126 while “brain fog” may evolve from low risk of severe disease, assessed 141 days following PTSD or deconditioning following critical illness and infection, had mild impairment of the pancreas. This invasive treatment.127 Evidence suggests that a direct impairment was associated with diarrhea, fever, viral encephalitis, systemic inflammation, peripheral headache, and dyspnea.13 Postmortem and case organ dysfunction, and cerebrovascular changes may studies have highlighted the impact that covid-19 contribute to the development of long term sequalae has on the spleen, including atrophy of lymphoid following covid-19.128 Figure 2 outlines the potential follicles, a decrease in T and B lymphocytes leading mechanisms occurring within the CNS. to lymphocytopenia, and thrombotic events such as infarcts.139-141 A cross sectional study found mild Olfactory and gustatory dysfunction impairment of the spleen in 4% of those assessed at Abnormalities of smell and taste have been reported 141 days following clearance of covid-19.13 Other to persist following recovery from covid-19. The ONS organs and tissues, such as the liver, gastrointestinal estimated the 5 week prevalence of loss of smell and tract, muscle, and blood vessels express the ACE2 loss of taste as 7.9% and 8.2% of all people who receptor and are susceptible to direct damage from have had covid-19, respectively.53 Other studies SARS-CoV-2 and indirect damage through elevated have found varying prevalence of olfactory and systemic inflammation.142-144 Alterations in gut http://www.bmj.com/ gustatory dysfunction, ranging from 11% to 45.1% microbiota145 and subacute thyroiditis142 have been of cohorts of patients who have recovered from acute observed following covid-19 infection. covid-19.22 39 47 Possible mechanisms Possible mechanisms Kidney injury may occur through several mechanisms Non-neuronal expression of the ACE2 receptor may associated with covid-19, including sepsis143 and

enable entry of the SARS-CoV-2 virus into olfactory lung injury leading to hemodynamic changes and on 11 August 2021 at Atlantic Health Hospitals. Protected by copyright. support cells, stem cells, and perivascular cells. This hypoxemia.144 The ACE2 receptor is highly expressed local infection could cause an inflammatory response in the pancreas,4 perhaps to a greater level than which subsequently reduces the function of olfactory in the lungs138; however, it is unclear whether sensory neurons. Additionally, by damaging the pancreatic damage is a direct result of viral infection support cells responsible for local water and ionic within the pancreas, or caused by the systemic balance, SARS-CoV-2 may indirectly reduce signaling inflammatory response seen during covid-19.146 The from sensory neurons to the brain,129 resulting in a spleen also expresses ACE2 receptors6 and may be loss of sense of smell. directly attacked by the virus, rather than the intense ACE2 receptors are also expressed on the mucous systemic inflammation being the primary cause of membrane of the oral cavity, particularly on the splenic damage.139 Chronic systemic inflammation tongue,130 therefore SARS-CoV-2 has a direct is frequently observed long after the clearance of route of entry into oral tissue, which may result in acute covid-19 infection,13 therefore, it is likely that cellular injury and dysfunction. Moreover, SARS- this elevated inflammatory state causes long term CoV-2 may bind to sialic acid receptors,131 causing complications in multiple organs in people with long an increase in gustatory threshold and resulting in covid. degradation of gustatory particles before they can be detected.132 Another possible mechanism of Risk factors gustatory dysfunction in covid-19 and long covid Risk factors for severe covid-19 and hospital concerns the functional link between taste and smell, admission, and risk factors for death as a whereby gustatory perception is reduced because of result of covid-19 include older age, male antecedent olfactory sensory dysfunction.133 sex, non-white ethnicity, being disabled, and pre-existing comorbidities including obesity, Other commonly reported manifestations cardiovascular disease, respiratory disease, and Covid-19 infection can result in multi-organ hypertension.2 13 19 147 148 Linked to risk of covid-19 impairment in individuals with low or high risk for severity and possibly the risk of long covid, the role of severe acute disease.2 13 Studies show the presence immune suppression is still being debated. Immune of acute kidney injury in discharged patients who suppression may have protective effects against long the bmj | BMJ 2021;374:n1648 | doi: 10.1136/bmj.n1648 9 STATE OF THE ART REVIEW BMJ: first published as 10.1136/bmj.n1648 on 26 July 2021. Downloaded from

term effects of covid-19 infection149-151; however, covid, suggested that the condition may actually these findings are conflicted.152 153 be four different syndromes.102 Recognizing which The risk factors for developing long covid are less patients belong to which subgroup of long covid, appreciated. To explore the characteristics associated and understanding the pathophysiology, will be with symptoms of long covid, 274 non-hospitalized important in deciding the treatment they receive. patients who had covid-19 were interviewed between 14 and 21 days following their positive test. Risk Guidelines factors for not returning to “usual health” included Various guidelines focus on treating and managing age (P=0.01), with the ≥50 years age group having long covid, or have included recommendations for the greatest odds ratio, and number of pre-existing long covid in their guidelines for treating covid-19.9 medical conditions (P=0.003), with a greater Guidelines recommend how to identify, refer, number of conditions associated with a greater odds and treat patients with long covid. The holistic ratio of not returning to “usual health.” Of the pre- assessment, investigation, and management existing conditions, having hypertension (odds ratio approaches suggested by NICE9 are outlined in (OR)=1.3, P=0.018), obesity (OR=2.31, P=0.002), fig 3. In January 2021, WHO updated its covid-19 a psychiatric condition (OR=2.32, P=0.007), or an guidance to include a new chapter focused on caring immunosuppressive condition (OR=2.33, P=0.047) for patients post-covid-19.157 These guidelines go corresponded with the greatest odds of not returning into little detail about long covid, however. Similarly, to “usual health.”18 the NIH has released treatment guidelines for A cross sectional study identified an association covid-19,158 but little guidance on managing long between the severity of acute covid-19 infection and covid. The CDC is expected to release guidance on post-recovery manifestations in people who have long covid management soon.159 The European had covid-19, showing that a more severe acute Society of Cardiology has also released guidelines phase may transform into the development of more on the diagnosis and management of cardiovascular severe symptoms of long covid.43 A cohort study, disease during the pandemic.160 The guidelines for meanwhile, corroborated this finding, with patients treating and managing long covid will undoubtedly with more than five symptoms during the initial evolve as new evidence comes to light; however, covid-19 infection and those that required hospital other general guidelines, such as Evidence Based http://www.bmj.com/ admission more likely to experience long covid Medicine’s guidance on post-infectious syndromes symptoms.34 may be useful for treating long covid.161 Although certain factors may increase the risk of both severe covid-19 and long covid, some factors Pulmonary symptoms associated with covid-19 do not also increase risk Pulmonary symptoms are common during long for long covid. Male sex and older age are associated covid. NICE recommends that breathlessness may be

with an increased risk of severe covid-19, however, investigated using an exercise tolerance test suited to on 11 August 2021 at Atlantic Health Hospitals. Protected by copyright. the ONS reported that the prevalence of any long the person’s ability, for example the one minute sit- covid symptoms is higher in women compared to-stand test, and treatment and management should with men (23.6% versus 20.7%), while the age be multidisciplinary, with advice and education group estimated to be most greatly affected by long given on managing breathlessness. Furthermore, covid symptoms is 35-49 years (26.8%), followed the guidelines recommend offering patients with by 50-69 years (26.1%), and the ≥70 years group continuing respiratory symptoms a chest radiograph (18%).53 Furthermore, a prospective cohort study by 12 weeks after infection.9 Blood oxygen levels can assessing recovered patients found no baseline be monitored using a pulse oximeter. clinical features associated with the subsequent Recommendations from the Mayo Clinic suggest development of long covid symptoms.154 Male sex, that shortness of breath can be self-managed by age, and pre-existing conditions including obesity, limiting factors that exacerbate dyspnea, including diabetes, and cardiovascular disease have shown no stopping smoking, avoiding pollutants, avoiding association with the risk of developing long covid. extremes in temperature, and exercising,162 however, However, pre-existence of asthma has been found to chronic shortness of breath may require further be significantly associated with long covid.34 intervention. Recognized non-pharmacological strategies for managing dyspnea include breathing Treatment and management of long covid exercises,163 pulmonary rehabilitation,164 and WHO and the Long Covid Forum Group agree maintaining optimal body positioning for postural that research priorities for long covid include relief.165 Meanwhile, a systematic review has found improving clinical characterization and the research that oral opioids can be used to treat dyspnea,166 and development of therapeutics.155 156 Clinical therefore this class of drugs may prove useful for characterization of patients with long covid is treating the condition in people with long covid. essential to provide appropriate treatment options. Patients with pulmonary fibrosis resulting from Gaining an understanding of why certain disease covid-19 should be managed in accordance with phenotypes arise in different individuals is an NICE guidelines on idiopathic pulmonary fibrosis,167 important piece of the puzzle. A review, which while antifibrotic therapies may be advantageous.168 included perspectives from patients with long Exacerbations of bronchiectasis should be treated

10 doi: 10.1136/bmj.n1648 | BMJ 2021;374:n1648 | the bmj STATE OF THE ART REVIEW BMJ: first published as 10.1136/bmj.n1648 on 26 July 2021. Downloaded from

Covid-19 diagnosis • Suspected acute covid-19 • Acute covid-19 diagnosed with positive test with hospitalisation • Retrospective clinical covid-19 diagnosis as self managed initial reaction • Acute covid-19 diagnosed with positive test without hospitalisation

Identifying long covid • Identifying people with ongoing symptomatic covid-19 or • Clinical review at 6 weeks post covid-19 syndrome and advise/inform on recovery

Assessment of people with new or ongoing symptoms aer acute covid-19 • Use holistic, person centred approach • Discuss person’s experiences with empathy • Include comprehensive clinical history and appropriate examinations • Do not predict likelihood of long covid based on acute symptoms • Consider wide ranging, uctuating symptoms that can occur aer acute covid-19 • Consider gradual decline, frailty or dementia a sign • Discuss effect of post covid-19 symptoms on daily life and work/education of long covid and use cognitive screening tools

Investigations and referral • Refer urgently to relevant service if symptoms could be life threatening • Offer tests depending on individual needs (exercise tolerance test, • Offer tailored tests and investigations to determine exact cause blood pressure and heart rate recordings, chest radiograph, etc) Covid-19 rapid guideline: managing long term effects of covid-19 (NICE, RCGP, SIGN) • If other (non-covid) diagnosis is suspected, • Consider referral (urgently if required) for people with psychiatric symptoms offer investigations in line with guidance • Aer ruling out life threatening conditions consider referral to multidisciplinary • Offer blood tests assessment service from 4 weeks aer start of acute covid-19

Planning care • Aer assessment, use shared decision making to discuss and agree with • When discussing appropriate level of support: person what support and rehabilitation they need including advice on - Think about overall impact of symptoms on life self management and support/referral to agreed clinical pathway - Look at overall trajectory of symptoms including uctuations http://www.bmj.com/ Management • Self management and supported self management • Multidisciplinary rehabilitation - Give advice and information on self management - Assess physical, psychological and psychiatric aspects of rehabilitation - Explain that it is unknown if over-the-counter vitamins and supplements are helpful - Work with person to develop rehabilitation and manage plan - Support people in discussing returning to work/education - Encourage people to monitor progress • Support for older people and children Consider additional support and referral for specialist advice on 11 August 2021 at Atlantic Health Hospitals. Protected by copyright. Follow-up and monitoring • Agree how oen follow-up is needed and which professionals should be involved • Consider self monitoring at home and ensure people have • Offer people in-person or remote monitoring using shared decision making clear instructions and parameters for when to seek help • Tailor monitoring to people’s symptoms and discuss changes • Be alert to symptoms developing that could mean referral or investigation is needed

Sharing information and continuity of care • Ensure effective information sharing and integrated working by sharing • Include baseline and ongoing assessments in shared information promptly between services and through multidisciplinary meetings information including when discharging person from hospital • Give people copy of their care plans or records • Provide continuity of care with same health care professional or team as much as possible

Service organisation • Provide access to multidisciplinary services for accessing • Share knowledge, skills and training between services to physical and mental health and carrying out further tests help practitioner to provide assessments and interventions • Provide integrated, multidisciplinary rehabilitation services, based on local need • Agree local, integrated referral pathways and resources. Core team could include (not limited to): Occupational therapy, physiotherapy, clinical psychology and psychiatry, rehabilitation medicine

Fig 3 | Overview of the NICE rapid guideline: managing the long term effects of covid-19

the bmj | BMJ 2021;374:n1648 | doi: 10.1136/bmj.n1648 11 STATE OF THE ART REVIEW BMJ: first published as 10.1136/bmj.n1648 on 26 July 2021. Downloaded from

with antimicrobial prescribing,169 while non- effectiveness and show a high incidence of adverse antimicrobial therapies, including airway clearance, events. This re-analysis study states that if a trial of may be considered.170 Modified rehabilitation a drug or surgical procedure demonstrated similarly practices, including stretching, body rotations, high rates of adverse effects, then it would not be acupressure, and massage have shown beneficial accepted as a safe treatment option, therefore CBT long term effects on respiratory symptoms in mild should have to adhere to the same level of scrutiny.182 covid-19 patients in a small trial.171 Another management strategy for fatigue is pacing, whereby patients manage tasks and activities Cardiovascular symptoms to avoid over-exertion and exacerbating fatigue. The NICE guidelines on long covid state that exercise NICE guidelines for ME/CFS178 describe pacing as tolerance tests may be undertaken to measure heart a self-management strategy, however guidance function, while lying and standing blood pressure and education from healthcare professionals may and heart rate recordings should be performed if be useful for patients. Evidence from randomized postural orthostatic tachycardia syndrome (POTS) is controlled trials for the use of pacing in long covid suspected.9 Urgent referral should occur for people is yet to be seen. that have symptoms of a life threating complication, The implementation of group therapy via such as cardiac chest pain. videoconferencing in people with early psychosis The European Society of Cardiology has released during the covid-19 pandemic shows promising comprehensive guidance for the diagnosis and results, with a pilot study showing improvements in management of cardiovascular disease during the psychotic symptoms and self-esteem,183 however, a covid-19 pandemic.160 The range of cardiovascular review article provides information to suggest that conditions that can manifest in long covid translates CBT is ineffective in reducing long covid symptoms, to a wide range of potential therapeutic options, including fatigue, with only 10% of participants therefore, ongoing investigation and observation of achieving clinically meaningful improvements.184 cardiac biomarkers is important. NICE guidelines GET is a structured intervention plan consisting recommend β blockers for several cardiac complaints, of physical activities with a therapeutic goal.185 including angina,172 cardiac arrhythmias,173 and A systematic review of exercise therapy for CFS acute coronary syndromes,174 therefore, β blockers concluded that patients with ME/CFS generally feel http://www.bmj.com/ may be useful in the treatment of cardiovascular less fatigued and have improved sleep and physical manifestations of long covid. Myocarditis may resolve function following completion of exercise therapy, to naturally over time; however, supportive and/or a greater degree than following a program of either immunomodulating therapy may improve recovery, adaptive pacing or supportive listening.186 The NICE as a systematic review describes.175 A review has also guidelines on ME/CFS recommend GET; however, in suggested that anticoagulants may be used to reduce July 2020 NICE released a statement urging caution 176 the risks associated with hypercoagulability. when implementing GET for people recovering from on 11 August 2021 at Atlantic Health Hospitals. Protected by copyright. Meanwhile, advice and education, agents to maintain covid-19, stating that with guidelines currently being vascular tone, and agents to manage palpitations updated, these recommendations may change.187 have been shown by a randomized controlled trial This statement accompanies concerns over the and discussed in a review to be advantageous in the potential negative effects of GET, including post- treatment of POTS.89 177 exertional malaise.188 Evidence specific to covid-19 is lacking, therefore Treating fatigue, cognitive, and neuropsychiatric cognitive impairment should be managed with symptoms support, including setting tailored, achievable Chronic fatigue is a common manifestation of long goals and implementing validated screening tools.9 covid. NICE recommends that self-management Managing cognitive impairment will require a and support are important in managing fatigue, holistic approach, however, patients should be owing to the poor availability of covid-19 specific advised that most people gradually recover from treatment.9 A condition that may overlap with long cognitive impairment following severe illness.106 189 covid fatigue is myalgic encephalomyelitis/chronic The holistic approach to treatment should extend fatigue syndrome (ME/CFS), therefore, the treatment to the services offered, with professionals including algorithm designed for treating ME/CFS may prove occupational and speech and language therapists useful in treating post-covid-19 fatigue. NICE has addressing cognitive changes.190 Cognitive specific guidelines that outline how to refer and treat impairment in long covid, sometimes called “brain ME/CFS patients; these include cognitive behavioral fog,” has been compared to “chemobrain.”191 The therapy (CBT) and graded exercise therapy (GET).178 Mayo clinic recommendations suggest strategies to Following backlash over these guidelines from the manage chemobrain including repeating exercises, ME Association,179 however, NICE aims to publish tracking what influences deficits, and using revised guidelines in August 2021.180 stress relief and coping strategies. Furthermore, Randomized controlled trials have shown that CBT medications including methylphenidate, donepezil, is beneficial in the treatment of chronic fatigue,181 modafinil, and memantine may be considered.192 however, this is conflicted by findings from a re- These strategies may prove useful for long covid. analysis of a Cochrane review which question its Specific to long covid, luteolin, a natural flavonoid,

12 doi: 10.1136/bmj.n1648 | BMJ 2021;374:n1648 | the bmj STATE OF THE ART REVIEW BMJ: first published as 10.1136/bmj.n1648 on 26 July 2021. Downloaded from

may alleviate cognitive impairment by inhibiting A trial to assess the effectiveness of an 8 week mast cell and microglia activation,191 but clinical exercise program in patients with long covid and trials are required. fatigue is ongoing (NCT04841759). Vitamin C Sleep disturbances may be managed by following supplementation may prove useful in treating fatigue relevant guidelines on insomnia,193 and a range in patients with long covid, with a systematic review of treatment strategies can be considered.194-197 concluding that high dose intravenous vitamin C could Patients with mental health problems alongside or be a beneficial treatment option.211 LOVIT-COVID as a result of long covid can be managed following (NCT04401150) is an ongoing clinical trial aimed at the relevant guidelines: depression,198 anxiety,199 assessing the effects of high dose intravenous vitamin PTSD,200 obsessive-compulsive disorder,201 and other C on hospitalized patients with covid-19. mental health problems.202 Care home residents, Two trials examining the effects of nicotinamide including those with dementia, who acquire long riboside, a dietary supplement, are ongoing covid have additional needs.116 Discussing mental (NCT04809974, NCT04604704) with the expectation health problems with patients requires compassion that the molecule reduces cognitive symptoms and understanding.203 and fatigue by modulating the pro-inflammatory response.212 Treating other organ impairments A clinical trial is currently ongoing assessing the Current evidence for the recovery of renal function effectiveness of a probiotic supplement to normalize following covid-19 is lacking. Considering that the composition of the gut microbiome and reduce early and close follow-ups with nephrologists inflammation in long covid (NCT04813718). The have previously been beneficial,204 post-covid-19 understanding of long term sequalae of covid-19 patients with renal dysfunction may benefit from infection in the gastrointestinal tract will evolve, early and ongoing monitoring. Covid-19 can disrupt with studies currently ongoing (NCT04691895), and alter the microbiome of the gut, which may which will subsequently affect treatment. allow for opportunistic infections.145 Covid-19 Other potential treatments are molecules that associated destructive thyroiditis can result in suppress the intense inflammatory response seen in incident hyperthyroidism, which can be treated covid-19. Leronlimab is a monoclonal antibody that with corticosteroids.142 Overall, close follow-up of blocks the function of CCL-5. It has been shown to http://www.bmj.com/ patients with long covid and adequate investigative be effective and safe in HIV213 and reduces plasma procedures should be kept up to accurately diagnose interleukin-6 levels in covid-19.214 Clinical trials and treat specific symptoms. are ongoing to evaluate the efficacy of leronlimab post-covid-19 (NCT04343651, NCT04347239, Repurposing drugs for long covid NCT04678830). Another antibody treatment, Antihistamines have been implicated as a possible tocilizumab, blocks interleukin-6 receptors and

treatment for covid-19, with a study that employed has shown efficacy in a small trial of patients with on 11 August 2021 at Atlantic Health Hospitals. Protected by copyright. cellular experiments suggesting that histamine-1 covid-19 patients.215 Trials to explore the effects antagonists may be able to reduce the covid-19 of tocilizumab are ongoing (NCT04330638). The infection rate by inhibiting SARS-CoV-2 from entering anti-oxidative and anti-inflammatory function of ACE2 expressing cells.205 Systematic reviews and melatonin may also be useful in treating long covid.216 molecular studies have suggested that histamine-1 Lastly, adjuvant treatments, such as adaptogens, are and histamine-2 antagonists are viable candidates being explored for their effectiveness in treating long for further clinical trials in covid-19.206-208 It covid (NCT04795557). remains to be seen whether antihistamines have potential for treating long covid. Antidepressants Conclusion have been proposed to reduce the effects of long With many people having been infected and covid. Antidepressant use has been associated with continuing to be infected with covid-19, the long reduced risk of intubation or death in covid-19,209 term implications are of increasing concern. Here, while a meta-analysis of antidepressant drug we have reviewed the studies that have explored treatment for major depressive disorder has shown the persisting symptoms of long covid, and have that use of antidepressants, including serotonin- addressed the possible risk factors associated with norepinephrine reuptake inhibitors and selective developing long covid and the treatment options that serotonin reuptake inhibitors, results in a reduction may be useful in alleviating its symptoms. Currently, in peripheral inflammatory markers.210 long covid remains enigmatic and, with the question of the impact that new variants of covid-19 will Emerging treatments have on the incidence and severity of long covid still Clinical trials exploring the efficacy of hyperbaric oxy­ looming large, it is important that research continues gen (NCT04842448), montelukast (NCT04695704), to explore post-covid-19 syndrome. Greater and deupirfenidone (NCT04652518) to treat respira­ understanding of the pathogenesis, risk factors, tory conditions in long covid are ongoing. A trial of symptoms, and methods of treating long covid is breath­ing exercises and singing is also under way to required to reduce the strain and demand on people assess their utility in improving breathing abnormalities with the condition and the healthcare systems that in patients with long covid (NCT04810065). will endeavor to support them. the bmj | BMJ 2021;374:n1648 | doi: 10.1136/bmj.n1648 13 STATE OF THE ART REVIEW BMJ: first published as 10.1136/bmj.n1648 on 26 July 2021. Downloaded from

6 Hamming I, Timens W, Bulthuis ML, Lely AT, Navis G, van Goor H. RESEARCH QUESTIONS Tissue distribution of ACE2 protein, the functional receptor for SARS coronavirus. A first step in understanding SARS pathogenesis. J • What is the precise epidemiology of long covid Pathol 2004;203:631-7. doi:10.1002/path.1570 and how will novel variants of covid-19 affect the 7 Wu Z, McGoogan JM. Characteristics of and important lessons epidemiology and severity of long covid? from the coronavirus disease 2019 (COVID-19) outbreak in China: summary of a report of 72 314 cases from the Chinese Center • What are the major risk factors for long covid and how for Disease Control and Prevention. JAMA 2020;323:1239-42. do we best reduce an individual’s risk of developing doi:10.1001/jama.2020.2648 long term post-covid-19 symptoms? 8 Ladds E, Rushforth A, Wieringa S, et al. Persistent symptoms after Covid-19: qualitative study of 114 “long Covid” patients and draft • Which symptoms, or set of symptoms, can we use to quality principles for services. BMC Health Serv Res 2020;20:1144. classify long covid, clinically and phenotypically, with doi:10.1186/s12913-020-06001-y 9 National Institute for Health and Care Excellence. COVID-19 rapid the aim of improving diagnosis and management? guideline: managing the long-term effects of COVID-19 NICE • What is the optimal treatment and management guideline; c2020. https://www.nice.org.uk/guidance/ng188 strategy for long covid and is this strategy non- 10 Datta SD, Talwar A, Lee JT. A proposed framework and timeline of the spectrum of disease due to SARS-CoV-2 infection: specific or will it require targeting and tailoring to illness beyond acute infection and public health implications. specific patients? JAMA 2020;324:2251-2. doi:10.1001/jama.2020.22717 11 Raman B, Cassar MP, Tunnicliffe EM, et al. Medium-term effects of SARS-CoV-2 infection on multiple vital organs, exercise capacity, cognition, quality of life and mental health, post-hospital HOW PATIENTS WERE INVOLVED IN THE CREATION discharge. EClinicalMedicine 2021;31:100683. doi:10.1016/j. OF THIS ARTICLE eclinm.2020.100683 12 Sollini M, Ciccarelli M, Cecconi M, et al. Vasculitis changes in 18 Members of a long covid focus group were contacted COVID-19 survivors with persistent symptoms: an [ F]FDG-PET/CT study. Eur J Nucl Med Mol Imaging 2021;48:1460-6. doi:10.1007/ and requested to review the initial drafts of this s00259-020-05084-3 article. The feedback received assisted in developing 13 Dennis A, Wamil M, Kapur S, et al. Multi-organ impairment in low-risk and focusing our review towards the experiences of individuals with long COVID. MedRxiv 2020;10.14.20212555 [preprint] doi:10.1101/2020.10.14.20212555 different symptoms experienced by patients with long 14 Pan L, Mu M, Yang P, et al. Clinical characteristics of COVID-19 covid. Cognition and mental health were of particular patients with digestive symptoms in Hubei, China: a descriptive, interest to patients, which we have addressed in this cross-sectional, multicenter study. Am J Gastroenterol 2020;115:766-73. doi:10.14309/ajg.0000000000000620 article. 15 Carfì A, Bernabei R, Landi F, Gemelli Against COVID-19 Post-Acute Care Study Group. Gemelli against C-P-ACSG. Persistent symptoms in http://www.bmj.com/ patients after acute COVID-19. JAMA 2020;324:603-5. doi:10.1001/ Contributors: HC, SR, JN, and MY performed the primary literature jama.2020.12603 search and drafted parts of the manuscript; HC was the first author 16 Mandal S, Barnett J, Brill SE, et al, ARC Study Group. ‘Long-COVID’: of the manuscript who drafted the manuscript and revised it; PE was a cross-sectional study of persisting symptoms, biomarker and responsible for the concept and design of the work. PE reviewed and imaging abnormalities following hospitalisation for COVID-19. revised the manuscript. PE is the guarantor. Thorax 2021;76:396-8. doi:10.1136/thoraxjnl-2020-215818 17 Halpin SJ, McIvor C, Whyatt G, et al. Postdischarge symptoms and Competing interests: We have read and understood the BMJ policy rehabilitation needs in survivors of COVID-19 infection: A cross- on declaration of interests and declare the following interests: PE sectional evaluation. J Med Virol 2021;93:1013-22. doi:10.1002/ was funded by the Medical Research Council and now by Higher jmv.26368 on 11 August 2021 at Atlantic Health Hospitals. Protected by copyright. Education Funding Council for England (HEFCE). He has also received 18 Tenforde MW, Kim SS, Lindsell CJ, et al, IVY Network Investigators, grants from Alzheimer’s Research, UK, Alzheimer’s Drug Discovery CDC COVID-19 Response Team, IVY Network Investigators. Foundation, Alzheimer’s Society, UK, Medical Research Council, Symptom duration and risk factors for delayed return to usual Alzheimer’s Association US, Van-Geest foundation, and European health among outpatients with COVID-19 in a multistate health care Union grants. PE is a consultant to Roche, Pfizer, and Novo Nordisk. systems network—United States, March-June 2020. MMWR Morb He has received educational and research grants from GE Healthcare, Mortal Wkly Rep 2020;69:993-8. doi:10.15585/mmwr.mm6930e1 Novo Nordisk, Piramal Life Science/Life Molecular Imaging, Avid 19 Goërtz YMJ, Van Herck M, Delbressine JM, et al. Persistent Radiopharmaceuticals and Eli Lilly. He is a member of the Scientific symptoms 3 months after a SARS-CoV-2 infection: the Advisory Board at Novo Nordisk. post-COVID-19 syndrome?ERJ Open Res 2020;6:00542. doi:10.1183/23120541.00542-2020 Provenance and peer review: commissioned; externally peer 20 Townsend L, Dyer AH, Jones K, et al. Persistent fatigue following reviewed. SARS-CoV-2 infection is common and independent of severity of This article is made freely available for use in accordance with initial infection. PLoS One 2020;15:e0240784. doi:10.1371/ BMJ’s website terms and conditions for the duration of the covid-19 journal.pone.0240784 pandemic or until otherwise determined by BMJ. You may use, 21 Boscolo-Rizzo P, Borsetto D, Fabbris C, et al. Evolution of altered download and print the article for any lawful, non-commercial purpose sense of smell or taste in patients with mildly symptomatic COVID-19. (including text and data mining) provided that all copyright notices JAMA Otolaryngol Head Neck Surg 2020;146:729-32. doi:10.1001/ jamaoto.2020.1379 and trade marks are retained. 22 Paderno A, Mattavelli D, Rampinelli V, et al. Olfactory and gustatory outcomes in COVID-19: a prospective evaluation in nonhospitalized 1 World Health Organization. Weekly operational update on subjects. Otolaryngol Head Neck Surg 2020;163:1144-9. COVID-19—4 July 2021. Weekly operational update on COVID-19 - 5 doi:10.1177/0194599820939538 July 2021 (who.int) 23 Puntmann VO, Carerj ML, Wieters I, et al. Outcomes of cardiovascular 2 Chen T, Wu D, Chen H, et al. Clinical characteristics of 113 deceased magnetic resonance imaging in patients recently recovered from patients with coronavirus disease 2019: retrospective study. coronavirus disease 2019 (COVID-19). JAMA Cardiol 2020;5:1265- BMJ 2020;368:m1091. doi:10.1136/bmj.m1091 73. doi:10.1001/jamacardio.2020.3557 3 Kim GU, Kim MJ, et al. Clinical characteristics of asymptomatic and 24 Helms J, Kremer S, Merdji H, et al. Neurologic features in severe SARS- symptomatic patients with mild covid-19. Clin Microbiol Infect CoV-2 infection. N Engl J Med 2020;382:2268-70. doi:10.1056/ 2020;26:948e1-3. NEJMc2008597 4 Hoffmann M, Kleine-Weber H, Schroeder S, et al. SARS-CoV-2 25 Vaes AW, Machado FVC, Meys R, et al. Care dependency in non- Cell entry depends on ACE2 and TMPRSS2 and is blocked by a hospitalized patients with COVID-19. J Clin Med 2020;9:2946. clinically proven protease inhibitor. Cell 2020;181:271-280.e8. doi:10.3390/jcm9092946 doi:10.1016/j.cell.2020.02.052 26 Arnold DT, Hamilton FW, Milne A, et al. Patient outcomes after 5 Hussman JP. Cellular and molecular pathways of COVID-19 hospitalisation with COVID-19 and implications for follow-up: and potential points of therapeutic intervention. Front results from a prospective UK cohort. Thorax 2021;76:399-401. Pharmacol 2020;11:1169. doi:10.3389/fphar.2020.01169 doi:10.1136/thoraxjnl-2020-216086

14 doi: 10.1136/bmj.n1648 | BMJ 2021;374:n1648 | the bmj STATE OF THE ART REVIEW BMJ: first published as 10.1136/bmj.n1648 on 26 July 2021. Downloaded from

27 D’ Cruz RF, Waller MD, Perrin F, et al Chest radiography is a poor 51 Chopra V, Flanders SA, O’Malley M, Malani AN, Prescott HC. Sixty-day predictor of respiratory symptoms and functional impairment outcomes among patients hospitalized with COVID-19. Ann Intern in survivors of severe COVID-19 pneumonia. ERJ Open Med 2021;174:576-8. doi:10.7326/M20-5661 Res 2020;7:00655. 52 Davis HE, Assaf GS, McCorkell L, et al. Characterizing long COVID in 28 Daher A, Balfanz P, Cornelissen C, et al. Follow up of patients with an international cohort: 7 months of symptoms and their impact. severe coronavirus disease 2019 (COVID-19): Pulmonary and MedRxiv 2020;12.24.20248802. extrapulmonary disease sequelae. Respir Med 2020;174:106197. 53 UK Office for National Statistics. Prevalence of long COVID doi:10.1016/j.rmed.2020.106197 symptoms and COVID-19 complications. 2020. https:// 29 Huang L, Zhao P, Tang D, et al. Cardiac involvement in patients www.ons.gov.uk/peoplepopulationandcommunity/ recovered from COVID-2019 identified using magnetic resonance healthandsocialcare/healthandlifeexpectancies/datasets/ imaging. JACC Cardiovasc Imaging 2020;13:2330-9. doi:10.1016/j. prevalenceoflongcovidsymptomsandcovid19complications jcmg.2020.05.004 54 Davies NG, Abbott S, Barnard RC, et al, CMMID COVID-19 Working 30 Huang Y, Tan C, Wu J, et al. Impact of coronavirus disease 2019 Group, COVID-19 Genomics UK (COG-UK) Consortium. Estimated on pulmonary function in early convalescence phase. Respir transmissibility and impact of SARS-CoV-2 lineage B.1.1.7 in England. Res 2020;21:163. doi:10.1186/s12931-020-01429-6 Science 2021;372:eabg3055. doi:10.1126/science.abg3055 31 Savarraj JPJ, Burkett AB, Hinds SN, et al. Three-month outcomes 55 UK Government. NERVTAG: update note on B.1.1.7 severity, 11 in hospitalized COVID-19 patients. MedRxiv 2020 [preprint] February 2021. https://www.gov.uk/government/publications/ 10.16.20211029. nervtag-update-note-on-b117-severity-11-february-2021 32 Sonnweber T, Boehm A, Sahanic S, et al. Persisting alterations of 56 UK Government. Variants: distribution of cases data. 2021. iron homeostasis in COVID-19 are associated with non-resolving https://www.gov.uk/government/publications/covid-19-variants- lung pathologies and poor patients’ performance: a prospective genomically-confirmed-case-numbers/variants-distribution-of-cases- observational cohort study. Respir Res 2020;21:276. doi:10.1186/ data s12931-020-01546-2 57 cov-lineages.org. B.1.1.7. 2021https://cov-lineages.org/global_ 33 Valiente-De S, Pérez-Camacho I, Sobrino B, et al. Clinical and report_B.1.1.7.html immunoserological status 12 weeks after infection with COVID-19: 58 Centers for Disease Control and Prevention. SARS-CoV-2 variant prospective observational study. MedRxiv [preprint] 2020. classifications and definitions. 2021. https://www.cdc.gov/ 2020.10.06.20206060. coronavirus/2019-ncov/cases-updates/variant-surveillance/variant- 34 Sudre CH, Murray B, Varsavsky T, et al. Attributes and predictors of info.html long COVID. Nat Med 2021;27:626-31. doi:10.1038/s41591-021- 59 Garner P. BMJ Opinion. Paul Garner: For 7 weeks I have been through 01292-y a roller coaster of ill health, extreme emotions, and utter exhaustion. 35 Vaira LA, Hopkins C, Petrocelli M, et al. Smell and taste recovery 2020. https://blogs.bmj.com/bmj/2020/05/05/paul-garner-people- in coronavirus disease 2019 patients: a 60-day objective and who-have-a-more-protracted-illness-need-help-to-understand-and- prospective study. J Laryngol Otol 2020;134:703-9. doi:10.1017/ cope-with-the-constantly-shifting-bizarre-symptoms/ S0022215120001826 60 Perego E, Callard F, Stras L, Melville-Jóhannesson B, Pope R, Alwan 36 Tomasoni D, Bai F, Castoldi R, et al. Anxiety and depression symptoms NA. Why the patient-made term ‘long covid’ is needed.Wellcome after virological clearance of COVID-19: A cross-sectional study in Open Research . 2020;5. doi:10.12688/wellcomeopenres.16307.1. Milan, Italy. J Med Virol 2021;93:1175-9. doi:10.1002/jmv.26459 61 Tansey CM, Louie M, Loeb M, et al. One-year outcomes and 37 Mazza MG, De Lorenzo R, Conte C, et al, COVID-19 BioB Outpatient health care utilization in survivors of severe acute respiratory http://www.bmj.com/ Clinic Study group. Anxiety and depression in COVID-19 survivors: syndrome. Arch Intern Med 2007;167:1312-20. doi:10.1001/ Role of inflammatory and clinical predictors. Brain Behav archinte.167.12.1312 Immun 2020;89:594-600. doi:10.1016/j.bbi.2020.07.037 62 Islam MF, Cotler J, Jason LA. Post-viral fatigue and COVID-19: lessons 38 Klein H, Asseo K, Karni N, et al. Onset, duration, and persistence from past epidemics. Fatigue 2020;8:61-9. doi:10.1080/21641846 of taste and smell changes and other COVID-19 symptoms: .2020.1778227. longitudinal study in Israeli patients. MedRxiv [preprint] 2020. 63 Wostyn P. COVID-19 and chronic fatigue syndrome: Is the worst 2020.09.25.20201343. doi:10.1101/2020.09.25.20201343. yet to come?Med Hypotheses 2021;146:110469. doi:10.1016/j. 39 Fjaeldstad AW. Prolonged complaints of chemosensory loss after mehy.2020.110469 COVID-19. Dan Med J 2020;67:A05200340. 64 Guedj E, Million M, Dudouet P, et al. 18F-FDG brain PET 40 Eiros R, Barreiro-Perez M, Martin-Garcia A, et al. Pericarditis and hypometabolism in post-SARS-CoV-2 infection: substrate on 11 August 2021 at Atlantic Health Hospitals. Protected by copyright. myocarditis long after SARS-CoV-2 infection: a cross-sectional for persistent/delayed disorders?Eur J Nucl Med Mol descriptive study in health-care workers. MedRxiv [preprint] 2020. Imaging 2021;48:592-5. doi:10.1007/s00259-020-04973-x 2020.07.12.20151316. 65 Delorme C, Paccoud O, Kas A, et al, CoCo-Neurosciences study group 41 Xiong Q, Xu M, Li J, et al. Clinical sequelae of COVID-19 survivors and COVID SMIT PSL study group. COVID-19-related encephalopathy: in Wuhan, China: a single-centre longitudinal study. Clin Microbiol a case series with brain FDG-positron-emission tomography/ Infect 2021;27:89-95. doi:10.1016/j.cmi.2020.09.023 computed tomography findings. Eur J Neurol 2020;27:2651-7. 42 Weerahandi H, Hochman KA, Simon E, et al. Post-discharge health doi:10.1111/ene.14478 status and symptoms in patients with severe COVID-19. J Gen Intern 66 Morgul E, Bener A, Atak M, et al. COVID-19 pandemic and Med 2021;36:738-45. doi:10.1007/s11606-020-06338-4 psychological fatigue in Turkey. Int J Soc Psychiatry 2021;67:128-35. 43 Kamal M, Abo Omirah M, Hussein A, Saeed H. Assessment and doi:10.1177/0020764020941889 characterisation of post-COVID-19 manifestations. Int J Clin 67 Brooks SK, Webster RK, Smith LE, et al. The psychological impact Pract 2021;75:e13746. doi:10.1111/ijcp.13746 of quarantine and how to reduce it: rapid review of the evidence. 44 Poyraz BÇ, Poyraz CA, Olğun Y, et al. Psychiatric morbidity Lancet 2020;395:912-20. doi:10.1016/S0140-6736(20)30460-8 and protracted symptoms after COVID-19. Psychiatry 68 Ferrandi PJ, Alway SE, Mohamed JS. The interaction between SARS- Res 2021;295:113604. doi:10.1016/j.psychres.2020.113604 CoV-2 and ACE2 may have consequences for skeletal muscle viral 45 Landi F, Carfì A, Benvenuto F, et al, Gemelli Against COVID-19 susceptibility and myopathies. J Appl Physiol (1985) 2020;129:864- Post-Acute Care Team. Predictive factors for a new positive 7. doi:10.1152/japplphysiol.00321.2020 nasopharyngeal swab among patients recovered from COVID-19. Am 69 Jin M, Tong Q. Rhabdomyolysis as potential late complication J Prev Med 2021;60:13-9. doi:10.1016/j.amepre.2020.08.014 associated with COVID-19. Emerg Infect Dis 2020;26:1618-20. 46 Carvalho-Schneider C, Laurent E, Lemaignen A, et al. Follow-up doi:10.3201/eid2607.200445 of adults with noncritical COVID-19 two months after symptom 70 Arnold P, Njemini R, Vantieghem S, et al. Peripheral muscle fatigue in onset. Clin Microbiol Infect 2021;27:258-63. doi:10.1016/j. hospitalised geriatric patients is associated with circulating markers cmi.2020.09.052 of inflammation. Exp Gerontol 2017;95:128-35. doi:10.1016/j. 47 Otte MS, Eckel HNC, Poluschkin L, Klussmann JP, Luers JC. Olfactory exger.2017.05.007 dysfunction in patients after recovering from COVID-19.Acta Otolaryngol 71 Chaudhuri A, Behan PO. Fatigue in neurological disorders. 2020;140:1032-5. doi:10.1080/00016489.2020.1811999 Lancet 2004;363:978-88. doi:10.1016/S0140-6736(04)15794-2 48 Zhao YM, Shang YM, Song WB, et al. Follow-up study of the 72 Wong TL, Weitzer DJ. Long COVID and myalgic encephalomyelitis/ pulmonary function and related physiological characteristics of chronic fatigue syndrome (ME/CFS)-a systemic review and COVID-19 survivors three months after recovery. EClinicalMedicine comparison of clinical presentation and symptomatology. Medicina 2020;25:100463. doi:10.1016/j.eclinm.2020.100463 (Kaunas) 2021;57:418. doi:10.3390/medicina57050418 49 Frontera JA, Yang D, Lewis A, et al. A prospective study of long-term 73 Mo X, Jian W, Su Z, et al. Abnormal pulmonary function in outcomes among hospitalized COVID-19 patients with and without COVID-19 patients at time of hospital discharge. Eur Respir neurological complications. J Neurol Sci 2021;426:117486. J 2020;55:2001217. doi:10.1183/13993003.01217-2020 doi:10.1016/j.jns.2021.117486 74 Kempuraj D, Selvakumar GP, Ahmed ME, et al. Covid-19, 50 Huang C, Huang L, Wang Y, et al. 6-month consequences of mast cells, cytokine storm, psychological stress, and COVID-19 in patients discharged from hospital: a cohort study. neuroinflammation. Neuroscientist 2020;26:402-14. Lancet 2021;397:220-32. doi:10.1016/S0140-6736(20)32656-8 doi:10.1177/1073858420941476 the bmj | BMJ 2021;374:n1648 | doi: 10.1136/bmj.n1648 15 STATE OF THE ART REVIEW BMJ: first published as 10.1136/bmj.n1648 on 26 July 2021. Downloaded from

75 Ackermann M, Verleden SE, Kuehnel M, et al. Pulmonary vascular 100 Varatharaj A, Thomas N, Ellul MA, et al, CoroNerve Study endothelialitis, thrombosis, and angiogenesis in covid-19. N Engl J Group. Neurological and neuropsychiatric complications of Med 2020;383:120-8. doi:10.1056/NEJMoa2015432 COVID-19 in 153 patients: a UK-wide surveillance study. Lancet 76 Wei J, Yang H, Lei P, et al. Analysis of thin-section CT in patients with Psychiatry 2020;7:875-82. doi:10.1016/S2215-0366(20)30287-X coronavirus disease (COVID-19) after hospital discharge. J Xray Sci 101 Pinna P, Grewal P, Hall JP, et al. Neurological manifestations and Technol 2020;28:383-9. doi:10.3233/XST-200685 COVID-19: Experiences from a tertiary care center at the Frontline. J 77 Liu M, Lv F, Huang Y, Xiao K. Follow-up study of the chest CT Neurol Sci 2020;415:116969. doi:10.1016/j.jns.2020.116969 characteristics of COVID-19 survivors seven months after 102 Maxwell E. National Institute for Health Research. Living with recovery. Front Med (Lausanne) 2021;8:636298. doi:10.3389/ Covid-19: a dynamic review of the evidence around ongoing covid-19 fmed.2021.636298 symptoms (often called Long Covid). 2020. https://evidence.nihr. 78 Han X, Fan Y, Alwalid O, et al. Six-month follow-up chest CT findings ac.uk/wp-content/uploads/2020/10/Living-with-Covid-Themed- after severe covid-19 pneumonia. Radiology 2021;299:E177-86. Review-October-2020.pdf doi:10.1148/radiol.2021203153 103 Sleat D, Wain R, Miller B. Long covid: reviewing the science and 79 McElvaney OJ, McEvoy NL, McElvaney OF, et al. Characterization of assessing the risk. Tony Blair Institute for Global Change. 2020. the inflammatory response to severe covid-19 illness. Am J Respir https://institute.global/sites/default/files/articles/Long-Covid- Crit Care Med 2020;202:812-21. doi:10.1164/rccm.202005- Reviewing-the-Science-and-Assessing-the-Risk.pdf 1583OC 104 The Royal Society. Long Covid: what is it, and what is needed? 2020. 80 Moodley YP, Scaffidi AK, Misso NL, et al. Fibroblasts isolated from https://royalsociety.org/-/media/policy/projects/set-c/set-c-long- normal lungs and those with idiopathic pulmonary fibrosis differ in covid.pdf?la=en-GB&hash=AD0672CAB24E1ECD14C2B1781A79 interleukin-6/gp130-mediated cell signaling and proliferation. Am J 3F25 Pathol 2003;163:345-54. doi:10.1016/S0002-9440(10)63658-9 105 Girard TD, Thompson JL, Pandharipande PP, et al. Clinical phenotypes 81 Cui S, Chen S, Li X, Liu S, Wang F. Prevalence of venous of delirium during critical illness and severity of subsequent long- thromboembolism in patients with severe novel coronavirus term cognitive impairment: a prospective cohort study. Lancet Respir pneumonia. J Thromb Haemost 2020;18:1421-4. doi:10.1111/ Med 2018;6:213-22. doi:10.1016/S2213-2600(18)30062-6 jth.14830 106 Pandharipande PP, Girard TD, Jackson JC, et al, BRAIN-ICU Study 82 Zhou F, Yu T, Du R, et al. Clinical course and risk factors for mortality Investigators. Long-term cognitive impairment after critical illness. N of adult inpatients with COVID-19 in Wuhan, China: a retrospective Engl J Med 2013;369:1306-16. doi:10.1056/NEJMoa1301372 cohort study. Lancet 2020;395:1054-62. doi:10.1016/S0140- 107 Mao L, Jin H, Wang M, et al. Neurologic manifestations of hospitalized 6736(20)30566-3 patients with coronavirus disease 2019 in Wuhan, China. JAMA 83 Shi S, Qin M, Shen B, et al. Association of cardiac injury with mortality Neurol 2020;77:683-90. doi:10.1001/jamaneurol.2020.1127 in hospitalized patients with COVID-19 in Wuhan, China. JAMA 108 Akhmerov A, Marbán E. COVID-19 and the heart. Circ Cardiol 2020;5:802-10. doi:10.1001/jamacardio.2020.0950 Res 2020;126:1443-55. doi:10.1161/CIRCRESAHA.120.317055 84 Rajpal S, Tong MS, Borchers J, et al. Cardiovascular magnetic 109 Caress JB, Castoro RJ, Simmons Z, et al. COVID-19-associated resonance findings in competitive athletes recovering from COVID-19 Guillain-Barré syndrome: The early pandemic experience. Muscle infection. JAMA Cardiol 2021;6:116-8. Nerve 2020;62:485-91. doi:10.1002/mus.27024 85 Johansson M, Ståhlberg M, Runold M, et al. Long-haul post-COVID-19 110 Heneka MT, Golenbock D, Latz E, Morgan D, Brown R. Immediate and symptoms presenting as a variant of postural orthostatic tachycardia long-term consequences of COVID-19 infections for the development syndrome: the Swedish experience. JACC Case Rep 2021;3:573-80. of neurological disease. Alzheimers Res Ther 2020;12:69. http://www.bmj.com/ doi:10.1016/j.jaccas.2021.01.009 doi:10.1186/s13195-020-00640-3 86 Kanjwal K, Jamal S, Kichloo A, Grubb BP. New-onset postural 111 Chamberlain SR, Grant JE, Trender W, Hellyer P, Hampshire A. Post- orthostatic tachycardia syndrome following coronavirus disease traumatic stress disorder symptoms in COVID-19 survivors: online 2019 infection. J Innov Card Rhythm Manag 2020;11:4302-4. population survey. BJPsych Open 2021;7:e47. doi:10.1192/bjo.2021.3 doi:10.19102/icrm.2020.111102 112 Taquet M, Luciano S, Geddes JR, Harrison PJ. Bidirectional 87 Miglis MG, Prieto T, Shaik R, Muppidi S, Sinn DI, Jaradeh S. A case associations between COVID-19 and psychiatric disorder: report of postural tachycardia syndrome after COVID-19. Clin Auton retrospective cohort studies of 62 354 COVID-19 cases in the Res 2020;30:449-51. doi:10.1007/s10286-020-00727-9 USA. Lancet Psychiatry 2021;8:130-40. doi:10.1016/S2215- 88 Umapathi T, Poh MQW, Fan BE, Li KFC, George J, Tan JYL. Acute 0366(20)30462-4 hyperhidrosis and postural tachycardia in a COVID-19 patient. Clin 113 Usher K, Bhullar N, Jackson D. Life in the pandemic: Social isolation and on 11 August 2021 at Atlantic Health Hospitals. Protected by copyright. Auton Res 2020;30:571-3. doi:10.1007/s10286-020-00733-x mental health. J Clin Nurs 2020;29:2756-7. doi:10.1111/jocn.15290 89 Dani M, Dirksen A, Taraborrelli P, et al. Autonomic dysfunction in ‘long 114 Creese B, Khan Z, Henley W, et al. Loneliness, physical activity, and COVID’: rationale, physiology and management strategies. Clin Med mental health during COVID-19: a longitudinal analysis of depression (Lond) 2021;21:e63-7. doi:10.7861/clinmed.2020-0896 and anxiety in adults over the age of 50 between 2015 and 2020. Int 90 Zheng YY, Ma YT, Zhang JY, Xie X. COVID-19 and the cardiovascular Psychogeriatr 2021;33:505-14. doi:10.1017/S1041610220004135 system. Nat Rev Cardiol 2020;17:259-60. doi:10.1038/s41569- 115 Velayudhan L, Aarsland D, Ballard C. Mental health of 020-0360-5 people living with dementia in care homes during COVID-19 91 Perez-Bermejo JA, Kang S, Rockwood SJ, et al. SARS-CoV-2 infection of pandemic. Int Psychogeriatr 2020;32:1253-4. doi:10.1017/ human iPSC-derived cardiac cells predicts novel cytopathic features S1041610220001088 in hearts of COVID-19 patients. Sci Transl Med 2021;13:eabf7872. 116 Simonetti A, Pais C, Jones M, et al. Neuropsychiatric symptoms doi:10.1126/scitranslmed.abf7872 in elderly with dementia during COVID-19 pandemic: definition, 92 Tavazzi G, Pellegrini C, Maurelli M, et al. Myocardial localization treatment, and future directions. Front Psychiatry 2020;11:579842. of coronavirus in COVID-19 cardiogenic shock. Eur J Heart doi:10.3389/fpsyt.2020.579842 Fail 2020;22:911-5. doi:10.1002/ejhf.1828 117 Manca R, De Marco M, Venneri A. The impact of COVID-19 infection 93 Adeghate EA, Eid N, Singh J. Mechanisms of COVID-19-induced heart and enforced prolonged social isolation on neuropsychiatric failure: a short review. Heart Fail Rev 2021;26:363-9. doi:10.1007/ symptoms in older adults with and without dementia: a review. Front s10741-020-10037-x Psychiatry 2020;11:585540. doi:10.3389/fpsyt.2020.585540 94 Agricola E, Beneduce A, Esposito A, et al. Heart and lung 118 Liu D, Baumeister RF, Veilleux JC, et al. Risk factors associated with multimodality imaging in COVID-19. JACC Cardiovasc mental illness in hospital discharged patients infected with COVID-19 Imaging 2020;13:1792-808. doi:10.1016/j.jcmg.2020.05.017 in Wuhan, China. Psychiatry Res 2020;292:113297. doi:10.1016/j. 95 Liu PP, Blet A, Smyth D, Li H. The science underlying COVID-19: psychres.2020.113297 implications for the cardiovascular system. Circulation 2020;142:68- 119 Jiang Z, Zhu P, Wang L, et al. Psychological distress and sleep quality 78. doi:10.1161/CIRCULATIONAHA.120.047549 of COVID-19 patients in Wuhan, a lockdown city as the epicenter 96 Siripanthong B, Nazarian S, Muser D, et al. Recognizing COVID- of COVID-19. J Psychiatr Res 2021;136:595-602. doi:10.1016/j. 19-related myocarditis: The possible pathophysiology and jpsychires.2020.10.034 proposed guideline for diagnosis and management. Heart 120 Lee SY, Song KJ, Lim CS, et al. Operation and management of Seoul Rhythm 2020;17:1463-71. doi:10.1016/j.hrthm.2020.05.001 Metropolitan City Community Treatment Center for mild condition 97 Shoenfeld Y, Ryabkova VA, Scheibenbogen C, et al. Complex COVID-19 patients. J Korean Med Sci 2020;35:e367. doi:10.3346/ syndromes of chronic pain, fatigue and cognitive impairment linked jkms.2020.35.e367 to autoimmune dysautonomia and small fiber neuropathy. Clin 121 Zhang Y, Zhang H, Ma X, Di Q. Mental health problems during the Immunol 2020;214:108384. doi:10.1016/j.clim.2020.108384 COVID-19 pandemics and the mitigation effects of exercise: a 98 Goldstein DS. The possible association between COVID-19 and longitudinal study of college students in China. Int J Environ Res postural tachycardia syndrome. Heart Rhythm 2021;18:508-9. Public Health 2020;17:3722. doi:10.3390/ijerph17103722 doi:10.1016/j.hrthm.2020.12.007 122 Grossman ES, Hoffman YSG, Palgi Y, Shrira A. COVID-19 related 99 Needham EJ, Chou SH, Coles AJ, Menon DK. Neurological implications loneliness and sleep problems in older adults: Worries and resilience of COVID-19 infections. Neurocrit Care 2020;32:667-71. as potential moderators. Pers Individ Dif 2021;168:110371. doi:10.1007/s12028-020-00978-4 doi:10.1016/j.paid.2020.110371

16 doi: 10.1136/bmj.n1648 | BMJ 2021;374:n1648 | the bmj STATE OF THE ART REVIEW BMJ: first published as 10.1136/bmj.n1648 on 26 July 2021. Downloaded from

123 Yachou Y, El Idrissi A, Belapasov V, Ait Benali S. Neuroinvasion, 145 Zuo T, Zhang F, Lui GCY, et al. Alterations in gut microbiota of patients neurotropic, and neuroinflammatory events of SARS-CoV-2: with COVID-19 during time of hospitalization. Gastroenterology understanding the neurological manifestations in COVID-19 2020;159:944-955.e8. doi:10.1053/j.gastro.2020.05.048 patients. Neurol Sci 2020;41:2657-69. doi:10.1007/s10072- 146 Zippi M, Hong W, Traversa G, et al. Involvement of the exocrine 020-04575-3 pancreas during COVID-19 infection and possible pathogenetic 124 Troyer EA, Kohn JN, Hong S. Are we facing a crashing wave hypothesis: a concise review. Infez Med 2020;28:507-15. of neuropsychiatric sequelae of COVID-19? Neuropsychiatric 147 Huang R, Zhu L, Xue L, et al. Clinical findings of patients with symptoms and potential immunologic mechanisms. Brain Behav coronavirus disease 2019 in Jiangsu province, China: A retrospective, Immun 2020;87:34-9. doi:10.1016/j.bbi.2020.04.027 multi-center study. PLoS Negl Trop Dis 2020;14:e0008280. 125 Romero-Sánchez CM, Díaz-Maroto I, Fernández-Díaz E, et al. doi:10.1371/journal.pntd.0008280 Neurologic manifestations in hospitalized patients with COVID-19: 148 Yang J, Zheng Y, Gou X, et al. Prevalence of comorbidities and its The ALBACOVID registry. Neurology 2020;95:e1060-70. effects in patients infected with SARS-CoV-2: a systematic review doi:10.1212/WNL.0000000000009937 and meta-analysis. Int J Infect Dis 2020;94:91-5. doi:10.1016/j. 126 South K, McCulloch L, McColl BW, Elkind MS, Allan SM, Smith CJ. ijid.2020.03.017 Preceding infection and risk of stroke: An old concept revived 149 SeyedAlinaghi S, Ghadimi M, Hajiabdolbaghi M, et al. Prevalence by the COVID-19 pandemic. Int J Stroke 2020;15:722-32. of COVID-19-like symptoms among people living with HIV, doi:10.1177/1747493020943815 and using antiretroviral therapy for prevention and treatment. 127 Kaseda ET, Levine AJ. Post-traumatic stress disorder: A Curr HIV Res 2020;18:373-80. doi:10.2174/157016 differential diagnostic consideration for COVID-19 survivors. Clin 2X18666200712175535 Neuropsychol 2020;34:1498-514. doi:10.1080/13854046.2020. 150 Del Amo J, Polo R, Moreno S, et al, The Spanish HIV/COVID-19 1811894 Collaboration. Incidence and severity of COVID-19 in HIV-positive 128 Nalbandian A, Sehgal K, Gupta A, et al. Post-acute COVID-19 persons receiving antiretroviral therapy: a cohort study. Ann Intern syndrome. Nat Med 2021;27:601-15. doi:10.1038/s41591-021- Med 2020;173:536-41. doi:10.7326/M20-3689 01283-z 151 Hu Y, Ma J, Huang H, Vermund SH. Coinfection with HIV and SARS- 129 Brann DH, Tsukahara T, Weinreb C, et al. Non-neuronal expression CoV-2 in Wuhan, China: a 12-person case series. J Acquir Immune of SARS-CoV-2 entry genes in the olfactory system suggests Defic Syndr 2020;85:1-5. doi:10.1097/QAI.0000000000002424 mechanisms underlying COVID-19-associated anosmia. Sci 152 Kowalska JD, Kase K, Vassilenko A, et al. The characteristics of HIV- Adv 2020;6:eabc5801. doi:10.1126/sciadv.abc5801 positive patients with mild/asymptomatic and moderate/severe 130 Xu H, Zhong L, Deng J, et al. High expression of ACE2 receptor course of COVID-19 disease-A report from Central and Eastern Europe. of 2019-nCoV on the epithelial cells of oral mucosa. Int J Oral Int J Infect Dis 2021;104:293-6. doi:10.1016/j.ijid.2020.12.026 Sci 2020;12:8. doi:10.1038/s41368-020-0074-x 153 Costenaro P, Minotti C, Barbieri E, Giaquinto C, Donà D. SARS-CoV-2 131 Milanetti M, Miotto M, Rienzo LD. Monti M, Gosti G, Ruocco G. infection in people living with HIV: a systematic review. Rev Med In-silico evidence for two receptors based strategy of SARS-CoV-2. Virol 2021;31:1-12. doi:10.1002/rmv.2155 bioRxiv [preprint] 2020:11107. 154 Moreno-Pérez O, Merino E, Leon-Ramirez JM, et al, COVID19-ALC 132 Pushpass RG, Pellicciotta N, Kelly C, Proctor G, Carpenter GH. research group. Post-acute COVID-19 syndrome. Incidence and risk Reduced salivary mucin binding and glycosylation in older adults factors: A Mediterranean cohort study. J Infect 2021;82:378-83. influences taste in an in vitro cell model. Nutrients 2019;11:2280. doi:10.1016/j.jinf.2021.01.004 doi:10.3390/nu11102280 155 Carson G, Long Covid Forum G, Long Covid Forum Group. Research http://www.bmj.com/ 133 Small DM, Prescott J. Odor/taste integration and the perception of priorities for Long Covid: refined through an international multi- flavor. Exp Brain Res 2005;166:345-57. doi:10.1007/s00221-005- stakeholder forum. BMC Med 2021;19:84. doi:10.1186/s12916- 2376-9 021-01947-0 134 Richardson S, Hirsch JS, Narasimhan M, et al, the Northwell COVID-19 156 World Health Organization. A coordinated global research roadmap. Research Consortium. Presenting characteristics, comorbidities, 2020. https://www.who.int/publications/m/item/a-coordinated- and outcomes among 5700 patients hospitalized with COVID-19 global-research-roadmap in the New York City area. JAMA 2020;323:2052-9. doi:10.1001/ 157 World Health Organization. COVID-19 clinical management: living jama.2020.6775 guidance. 2021. https://www.who.int/publications/i/item/WHO- 135 Hirsch JS, Ng JH, Ross DW, et al, Northwell COVID-19 Research 2019-nCoV-clinical-2021-1 Consortium, Northwell Nephrology COVID-19 Research Consortium. 158 National Institute of Health. Coronavirus disease 2019 on 11 August 2021 at Atlantic Health Hospitals. Protected by copyright. Acute kidney injury in patients hospitalized with COVID-19. Kidney (COVID-19) treatment guidelines. 2021. https://www. Int 2020;98:209-18. doi:10.1016/j.kint.2020.05.006 covid19treatmentguidelines.nih.gov/ 136 Pei G, Zhang Z, Peng J, et al. Renal involvement and early 159 USnews.com. CDC expected to release guidance on identifying, prognosis in patients with COVID-19 pneumonia. J Am Soc managing long COVID. 2021. https://www.usnews.com/news/health- Nephrol 2020;31:1157-65. doi:10.1681/ASN.2020030276 news/articles/2021-05-07/cdc-to-release-clinical-guidance-on- 137 Hadi A, Werge M, Kristiansen KT, et al. Coronavirus Disease-19 identifying-managing-long-covid (COVID-19) associated with severe acute pancreatitis: Case 160 Escardio.org. ESC guidance for the diagnosis and management of CV report on three family members. Pancreatology 2020;20:665-7. disease during the COVID-19 pandemic. 2020. https://www.escardio. doi:10.1016/j.pan.2020.04.021 org/Education/COVID-19-and-Cardiology/ESC-COVID-19-Guidance 138 Liu F, Long X, Zhang B, Zhang W, Chen X, Zhang Z. ACE2 expression in 161 Lauhio A. Postinfectious syndromes. EBM guidelines.com. 2017. pancreas may cause pancreatic damage after SARS-CoV-2 infection. https://www.ebm-guidelines.com/go/ebm/ebm01118.html Clin Gastroenterol Hepatol 2020;18:2128-2130.e2. doi:10.1016/j. 162 Mayoclinic.org. Shortness of breath. 2020. https://www.mayoclinic. cgh.2020.04.040 org/symptoms/shortness-of-breath/basics/when-to-see-doctor/sym- 139 Xu X, Chang XN, Pan HX, et al. [Pathological changes of the 20050890 spleen in ten patients with coronavirus disease 2019(COVID-19) 163 Holland AE, Hill CJ, Jones AY, McDonald CF. Breathing exercises for by postmortem needle autopsy]. Zhonghua Bing Li Xue Za chronic obstructive pulmonary disease. Cochrane Database Syst Zhi 2020;49:576-82. Rev 2012;10:CD008250. doi:10.1002/14651858.CD008250.pub2 140 Ihlow J, Michaelis E, Greuel S, et al. B cell depletion and signs of 164 McCarthy B, Casey D, Devane D, Murphy K, Murphy E, Lacasse sepsis-acquired immunodeficiency in bone marrow and spleen Y. Pulmonary rehabilitation for chronic obstructive pulmonary of COVID-19 deceased. Int J Infect Dis 2021;103:628-35. disease. Cochrane Database Syst Rev 2015;2:CD003793. doi:10.1016/j.ijid.2020.12.078 doi:10.1002/14651858.CD003793.pub3 141 Santos Leite Pessoa M, Franco Costa Lima C, Farias Pimentel AC, 165 O’Neill S, McCarthy DS. Postural relief of dyspnoea in severe chronic Godeiro Costa JC, Bezerra Holanda JL. Multisystemic infarctions airflow limitation: relationship to respiratory muscle strength. in COVID-19: focus on the spleen. Eur J Case Rep Intern Thorax 1983;38:595-600. doi:10.1136/thx.38.8.595 Med 2020;7:001747. 166 Jennings AL, Davies AN, Higgins JP, Gibbs JS, Broadley KE. A 142 Ruggeri RM, Campennì A, Siracusa M, Frazzetto G, Gullo D. Subacute systematic review of the use of opioids in the management of thyroiditis in a patient infected with SARS-COV-2: an endocrine dyspnoea. Thorax 2002;57:939-44. doi:10.1136/thorax.57.11.939 complication linked to the COVID-19 pandemic. Hormones 167 National Institute for Health and Care Excellence. Idiopathic (Athens) 2021;20:219-21. doi:10.1007/s42000-020-00230-w pulmonary fibrosis in adults: diagnosis and management clinical 143 Zarbock A, Gomez H, Kellum JA. Sepsis-induced acute kidney guideline. 2017. https://www.nice.org.uk/guidance/cg163 injury revisited: pathophysiology, prevention and future 168 Torrisi SE, Kahn N, Vancheri C, Kreuter M. Evolution and treatment therapies. Curr Opin Crit Care 2014;20:588-95. doi:10.1097/ of idiopathic pulmonary fibrosis. Presse Med 2020;49:104025. MCC.0000000000000153 doi:10.1016/j.lpm.2020.104025 144 Husain-Syed F, Slutsky AS, Ronco C. Lung-kidney cross-talk in the 169 National Institute for Health and Care Excellence. Bronchiectasis critically ill patient. Am J Respir Crit Care Med 2016;194:402-14. (non-cystic fibrosis), acute exacerbation: antimicrobial prescribing doi:10.1164/rccm.201602-0420CP NICE guideline. 2018. https://www.nice.org.uk/guidance/ng117

the bmj | BMJ 2021;374:n1648 | doi: 10.1136/bmj.n1648 17 STATE OF THE ART REVIEW BMJ: first published as 10.1136/bmj.n1648 on 26 July 2021. Downloaded from

170 Bell SC, Elborn JS, Byrnes CA. Bronchiectasis: Treatment 195 Xie Z, Chen F, Li WA, et al. A review of sleep disorders and melatonin. decisions for pulmonary exacerbations and their prevention. Neurol Res 2017;39:559-65. doi:10.1080/01616412.2017.1315864 Respirology 2018;23:1006-22. doi:10.1111/resp.13398 196 van Maanen A, Meijer AM, van der Heijden KB, Oort FJ. The effects 171 Zha L, Xu X, Wang D, Qiao G, Zhuang W, Huang S. Modified of light therapy on sleep problems: A systematic review and rehabilitation exercises for mild cases of COVID-19. Ann Palliat meta-analysis. Sleep Med Rev 2016;29:52-62. doi:10.1016/j. Med 2020;9:3100-6. doi:10.21037/apm-20-753 smrv.2015.08.009 172 National Institute for Health and Care Excellence. Stable angina: 197 Yang PY, Ho KH, Chen HC, Chien MY. Exercise training improves management clinical guideline. 2016. https://www.nice.org.uk/ sleep quality in middle-aged and older adults with sleep problems: Guidance/CG126 a systematic review. J Physiother 2012;58:157-63. doi:10.1016/ 173 National Institute for Health and Care Excellence. Atrial fibrillation: S1836-9553(12)70106-6 diagnosis and management NICE guideline. 2021. https://www.nice. 198 National Institute for Health and Care Excellence. Depression in org.uk/guidance/ng196 adults: recognition and management clinical guideline. 2009. 174 National Institute for Health and Care Excellence. Acute coronary https://www.nice.org.uk/guidance/CG90 syndromes NICE guideline. 2020. https://www.nice.org.uk/ 199 National Institute for Health and Care Excellence. Social anxiety guidance/ng185 disorder: recognition, assessment and treatment clinical guideline. 175 Sinagra G, Anzini M, Pereira NL, et al. Myocarditis in clinical practice. 2013. https://www.nice.org.uk/guidance/CG159 Mayo Clin Proc 2016;91:1256-66. doi:10.1016/j.mayocp.2016.05.013 200 National Institute for Health and Care Excellence. Post-traumatic 176 Onishi A, St Ange K, Dordick JS, Linhardt RJ. Heparin and anticoagulation. stress disorder NICE guideline. 2018. https://www.nice.org.uk/ Front Biosci (Landmark Ed) 2016;21:1372-92. doi:10.2741/4462 guidance/ng116 177 Taub PR, Zadourian A, Lo HC, Ormiston CK, Golshan S, Hsu JC. 201 Fineberg NA, Van Ameringen M, Drummond L, et al. How to manage Randomized trial of ivabradine in patients with hyperadrenergic obsessive-compulsive disorder (OCD) under COVID-19: A clinician’s postural orthostatic tachycardia syndrome. J Am Coll guide from the International College of Obsessive Compulsive Cardiol 2021;77:861-71. doi:10.1016/j.jacc.2020.12.029 Spectrum Disorders (ICOCS) and the Obsessive-Compulsive and 178 National Institute for Health and Care Excellence. Chronic fatigue Related Disorders Research Network (OCRN) of the European College syndrome/myalgic encephalomyelitis (or encephalopathy): diagnosis of Neuropsychopharmacology. Compr Psychiatry 2020;100:152174. and management clinical guideline. 2007. https://www.nice.org.uk/ doi:10.1016/j.comppsych.2020.152174 Guidance/CG53 202 National Institute for Health and Care Excellence. Common mental 179 ME Association. ME Association statement on the NICE clinical health problems: identification and pathways to care clinical guideline for ME/CFS and the NICE guideline for post/long covid-19. guideline. 2011. https://www.nice.org.uk/guidance/cg123 2020. https://meassociation.org.uk/2020/10/me-association- 203 Uzunova G, Pallanti S, Hollander E. Presentation and management statement-on-the-nice-clinical-guideline-for-me-cfs-and-the-nice- of anxiety in individuals with acute symptomatic or asymptomatic guideline-for-post-long-covid-19/ COVID-19 infection, and in the post-COVID-19 recovery phase. Int J 180 National Institute for Health and Care Excellence. Myalgic Psychiatry Clin Pract 2021;25:115-31. doi:10.1080/13651501.20 encephalomyelitis (or encephalopathy)/chronic fatigue syndrome: 21.1887264 diagnosis and management—in development [GID-NG10091]. 2021. 204 Meier P, Bonfils RM, Vogt B, Burnand B, Burnier M. Referral patterns https://www.nice.org.uk/guidance/indevelopment/gid-ng10091 and outcomes in noncritically ill patients with hospital-acquired acute 181 Adamson J, Ali S, Santhouse A, Wessely S, Chalder T. Cognitive kidney injury. Clin J Am Soc Nephrol 2011;6:2215-25. doi:10.2215/ behavioural therapy for chronic fatigue and chronic fatigue CJN.01880211 http://www.bmj.com/ syndrome: outcomes from a specialist clinic in the UK. J R Soc 205 Ge S, Wang X, Hou Y, Lv Y, Wang C, He H. Repositioning of

Med 2020;113:394-402. doi:10.1177/0141076820951545 histamine H1 receptor antagonist: Doxepin inhibits viropexis of 182 Vink M, Vink-Niese A. Cognitive behavioural therapy for myalgic SARS-CoV-2 Spike pseudovirus by blocking ACE2. Eur J Pharmacol encephalomyelitis/chronic fatigue syndrome is not effective. 2021;896:173897. doi:10.1016/j.ejphar.2021.173897 Re-analysis of a Cochrane review. Health Psychol Open 206 Demopoulos C, Antonopoulou S, Theoharides TC. COVID-19, 2019;6:2055102919840614. doi:10.1177/2055102919840614 microthromboses, inflammation, and platelet activating factor. 183 Lecomte T, Abdel-Baki A, Francoeur A, et al. Group therapy via Biofactors 2020;46:927-33. doi:10.1002/biof.1696 videoconferencing for individuals with early psychosis: A pilot study. 207 Chenchula S, Ray A, Sadasivam B. Famotidine repurposing for novel Early Interv Psychiatry 2020. PubMed doi:10.1111/eip.13099. corona virus disease of 2019: a systematic review. Drug Res 2021. 184 Vink M, Vink-Niese A. Could cognitive behavioural therapy be https://pesquisa.bvsalud.org/global-literature-on-novel-coronavirus- on 11 August 2021 at Atlantic Health Hospitals. Protected by copyright. an effective treatment for long covid and post covid-19 fatigue 2019-ncov/resource/en/covidwho-1146573. syndrome? Lessons from the Qure Study for Q-Fever Fatigue 208 Ishola AA, Joshi T, Abdulai SI, Tijjani H, Pundir H, Chandra S. Molecular Syndrome. Healthcare (Basel) 2020;8:552. doi:10.3390/ basis for the repurposing of histamine H2-receptor antagonist to treat healthcare8040552 COVID-19. J Biomol Struct Dyn 2021;Jan 25:1-18. 185 Chodzko-Zajko WJ, Proctor DN, Fiatarone Singh MA, et al, 209 Hoertel N, Sánchez-Rico M, Vernet R, et al, AP-HP / Universities / American College of Sports Medicine. American College of Sports INSERM COVID-19 Research Collaboration and AP-HP COVID CDR Medicine position stand. Exercise and physical activity for older Initiative. Association between antidepressant use and reduced risk adults. Med Sci Sports Exerc 2009;41:1510-30. doi:10.1249/ of intubation or death in hospitalized patients with COVID-19: results MSS.0b013e3181a0c95c from an observational study. Mol Psychiatry 2021. doi:10.1038/ 186 Larun L, Brurberg KG, Odgaard-Jensen J, Price JR. Exercise s41380-021-01021-4. therapy for chronic fatigue syndrome. Cochrane Database Syst 210 Köhler CA, Freitas TH, Stubbs B, et al. Peripheral alterations in Rev 2017;4:CD003200. cytokine and chemokine levels after antidepressant drug treatment 187 National Institute for Health and Care Excellence. Statement about for major depressive disorder: systematic review and meta-analysis. graded exercise therapy in the context of COVID-19. 2020. https:// Mol Neurobiol 2018;55:4195-206. www.nice.org.uk/guidance/gid-ng10091/documents/statement 211 Vollbracht C, Kraft K. Feasibility of vitamin C in the treatment of post 188 Torjesen I. NICE cautions against using graded exercise therapy for viral fatigue with focus on long COVID, based on a systematic review patients recovering from covid-19. BMJ 2020;370:m2933. of IV vitamin C on fatigue. Nutrients 2021;13:1154. doi:10.3390/ 189 Herridge MS, Moss M, Hough CL, et al. Recovery and outcomes nu13041154 after the acute respiratory distress syndrome (ARDS) in patients 212 Omran HM, Almaliki MS. Influence of NAD+ as an ageing-related and their family caregivers. Intensive Care Med 2016;42:725-38. immunomodulator on COVID 19 infection: A hypothesis. J Infect doi:10.1007/s00134-016-4321-8 Public Health 2020;13:1196-201. doi:10.1016/j.jiph.2020.06.004 190 Physiotherapy.ca. Rehabilitation for patients with COVID-19. 213 Dhody K, Pourhassan N, Kazempour K, et al. PRO 140, a monoclonal Guidance for occupational therapists, physical therapists, speech- antibody targeting CCR5, as a long-acting, single-agent maintenance language pathologists and assistants. 2020. https://physiotherapy. therapy for HIV-1 infection. HIV Clin Trials 2018;19:85-93. doi:10.10 ca/sites/default/files/final_rehabilitation_for_patients_with_ 80/15284336.2018.1452842 covid_19_apr-08-2020_clean_133pm.pdf 214 Patterson BK, Seethamraju H, Dhody K, et al. CCR5 inhibition in 191 Theoharides TC, Cholevas C, Polyzoidis K, Politis A. Long-COVID critical COVID-19 patients decreases inflammatory cytokines, syndrome-associated brain fog and chemofog: Luteolin to the rescue. increases CD8 T-cells, and decreases SARS-CoV2 RNA in plasma Biofactors 2021;47:232-41. doi:10.1002/biof.1726 by day 14. Int J Infect Dis 2021;103:25-32. doi:10.1016/j. 192 Mayo Clinic. Chemo brain. 2021. https://www.mayoclinic.org/diseases- ijid.2020.10.101 conditions/chemo-brain/diagnosis-treatment/drc-20351065 215 Xu X, Han M, Li T, et al. Effective treatment of severe 193 National Institute for Health and Care Excellence. Insomnia: how COVID-19 patients with tocilizumab. Proc Natl Acad Sci U S A should I assess a person with suspected insomnia? 2021. https:// 2020;117:10970-5. doi:10.1073/pnas.2005615117 cks.nice.org.uk/topics/insomnia/diagnosis/assessment/ 216 Bahrampour Juybari K, Pourhanifeh MH, Hosseinzadeh A, Hemati 194 National Institute for Health and Care Excellence. Guidance on the use K, Mehrzadi S. Melatonin potentials against viral infections of zaleplon, zolpidem and zopiclone for the short-term management including COVID-19: Current evidence and new findings. Virus of insomnia. 2004. https://www.nice.org.uk/Guidance/TA77 Res 2020;287:198108. doi:10.1016/j.virusres.2020.198108

No commercial reuse: See rights and reprints http://www.bmj.com/permissions Subscribe: http://www.bmj.com/subscribe CME KEITH W. MILLETTE, MD, FAAFP, RPh

A Step-by-Step Time-Saving Approach to Coding Office Visits

Follow these four steps to code quickly and accurately, while reducing the need to count up data points. © ROBERT NEUBECKER

he new rules for coding evaluation and management (E/M) office visits are a big improvement but still a lot to digest.1,2 To ease the transition, previous FPM articles have laid out the new American Medical Association/CPT medical deci- sion Tmaking guide3 and introduced doctor–friendly coding tem- plates (see “Countdown to the E/M Coding Changes,” FPM, September/October 2020, https://www.aafp.org/fpm/2020/0900/p29. html), explained how to quickly identify level 4 office visits (see ABOUT THE AUTHOR “Coding Level 4 Visits Using the New E/M Guidelines,” FPM, January/ Dr. Millette is a family physician practicing at Altru Family Medicine Center in Grand Forks, N.D. February 2021, https://www.aafp.org/fpm/2021/0100/p27.html), and He has been teaching coding to physicians for applied the new guidelines to common visit types (see “The 2021 almost 20 years. Author disclosure: no relevant Office Visit Coding Changes: Putting the Pieces together,”FPM , financial affiliations.

www.aafp.org/fpmDownloaded from www.aafp.org/fpm. Copyright © 2021 American Academy of Family Physicians. For the private,July/August noncommercial 2021 use| FPM of one| 21 individual user of the website. All other rights reserved. Contact [email protected] for questions and permission requests. November/December 2020, https://www. an office visit. That means the “history” and aafp.org/fpm/2020/1100/p6.html). “physical exam” components are no longer After several months of using the new needed for code selection, which simplifies coding rules, it has become clear that the things. But your patient note must still con- most difficult chore of coding office visits tain a “medically appropriate” history and now is assessing data to determine the physical. So continue to document what is level of medical decision making (MDM). needed for good medical care. New patient. A new patient is a patient who has not been seen by you or one of your partners in the same medical spe- It has become clear that the most cialty and the same group practice within difficult chore of coding office visits now the past three years. Total time and prolonged services. is assessing data to determine the level Total time includes all the time you spend on a visit on the day of the encounter of medical decision making. (before midnight). It includes your time before the visit reviewing the chart, your face-to-face time with the patient, and Analyzing each note for data points can be the time you spend after the visit finish- time-consuming and sometimes confusing. ing documentation, ordering or reviewing That being the case, it’s important to studies, refilling medications, making understand when you can avoid using data phone calls related to the visit, etc. It does for coding, and when you can’t. I’ve devel- not include your time spent performing oped a four-step process for this (see “A separately billed services such as wellness step-by-step timesaver” on page 25). visits or procedures. Total time visit level The goal of this article is to clarify the thresholds differ for new patients vs. estab- new coding rules and terminology and to lished patients. (See the total times in “The explain this step-by-step approach to help Rosetta Stone four-step template for coding clinicians code office visits more quickly, office visits” on page 24.) confidently, and correctly. The prolonged services code comes into play when total time exceeds the limits set OFFICE VISIT CODING RULES AND for level 5 visits by at least 15 minutes. (See TERMINOLOGY “Prolonged services guide” on page 25.) To make full use of the step-by-step pro- Medical decision making. MDM is made cess, we have to first understand the new up of three components: problems, data, rules, as well as coding terminology. Here is and risk. Each component has different a brief summary. levels, which correspond to levels of service Medically appropriate. Physicians and (low/limited = level 3, moderate = level 4, other qualified health care professionals and high/extensive = level 5). The highest may now solely use either total time or level reached by at least two out of the MDM to determine the level of service of three components determines the correct code for the level of service. MDM criteria is the same for new and established patients. KEY POINTS Problems addressed. This includes only • The new evaluation and management office visit coding rules have the problems you address at that specific simplified many things but are still a lot to digest, especially when it patient visit. It does not include all the comes to counting data. patient’s diagnoses and does not include problems that are exclusively managed by • There are different levels of data and different categories within another clinician. Problems addressed are each level, which can make using data to calculate the visit level separated into low-complexity problems time-consuming and confusing. (level 3), moderate-complexity problems • By calculating total time, and then moving on to assessing problems (level 4), and high-complexity problems and prescription drug management, most visits can be optimally (level 5). To code correctly, you need to coded without dealing with data at all. know the coding value of the problems you address. It is helpful to think of problems

22 | FPM | July/August 2021 www.aafp.org/fpm CODING APPROACH

WHAT LEVEL OF PROBLEM DID I ADDRESS?

Level 2 problem Level 3 problem Level 4 problem Level 5 problem 1 minor or self-limited 2 minor or self-limited 1 new problem with uncertain problem (e.g., simple rash). problems prognosis (e.g., breast lump) 1 stable chronic illness* 2 stable or 1 or more unstable 1 chronic illness with severe chronic illnesses exacerbation or progression 1 uncomplicated acute 1 acute illness with systemic 1 life-threatening acute or illness (e.g., urinary symptoms (e.g., pneumonia, chronic illness (e.g., heart attack, tract infection, sinusitis, colitis, pyelonephritis, or stroke, pulmonary embolism, or sore throat) pelvic inflammatory disease) acute kidney injury, or depression with suicidal ideation) 1 uncomplicated injury 1 acute complicated injury 1 life-threatening injury (e.g., simple ankle sprain) (e.g., head injury with brief loss of consciousness)

* Chronic illnesses include hypertension, hyperlipidemia, coronary artery disease, chronic obstructive pulmonary disease (COPD), asthma, diabetes, obesity, etc. in terms of levels of service (e.g., a sinus Data analyzed. For purposes of MDM, infection is usually a level 3 problem, and data is characterized as “limited” (level 3 pneumonia or uncontrolled diabetes are data), “moderate” (level 4 data), or “exten- usually level 4 problems). sive” (level 5 data). But each level of data is The simplest way to summarize prob- further split into Categories 1, 2, and 3. This lems is this: Life-threatening problems are can make calculating data complicated, level 5; acute or chronic illnesses or injuries confusing, and time-consuming. Here are are level 3 or 4 depending on how many the data components and terms you need there are, how stable they are, and how to know. complex they are; and if there’s just one Category 1 data includes the following: minor problem, it’s level 2. • The ordering or reviewing of each (For more specifics see “What level of unique test, i.e., a single lab test, panel, X-ray, problem did I address?”) electrocardiogram (ECG), or other study. Risk. Risk is also separated into “low” (level 3), “moderate” (level 4), and “high” (level 5) categories. Level 3 risk includes the use of over-the- Each level of data is further split into counter (OTC) medications. Categories 1, 2, and 3. This can make Level 4 risk includes the following: • Prescription drug management: start- calculating data complicated, confusing, ing, stopping, modifying, refilling, or decid- ing to continue a prescription medication and time-consuming. (and documenting your thought process), • Social determinants of health that limit diagnosis or treatment (this is when Ordering and reviewing the same lab test patients’ lack of finances, insurance, food, or study is worth one point, not two; a lab housing, etc., affects your ability to diag- panel (e.g., complete blood count or compre- nose, manage, and care for them as you hensive metabolic panel) is worth one point, normally would). • Reviewing a pertinent test or study Level 5 risk includes the following: done in the past at your own facility or • Decisions about hospitalization, another facility, • Decisions about emergency major • Reviewing prior external notes from surgery, each unique source, including records from • Drug therapy that requires intensive a clinician in a different specialty or from a toxicity monitoring, different group practice or facility as well • Decisions to not resuscitate or to de- as each separate health organization (e.g., escalate care because of poor prognosis. reviewing three notes from the Mayo Clinic www.aafp.org/fpm July/August 2021 | FPM | 23 is worth one point, not three, but reviewing Category 2 data includes the following: one note from Mayo and one from Johns • Using an independent historian (for Hopkins is worth a total of two points), level 3 data only), • Using an independent historian, which • Independent interpretation of tests, means obtaining a history from someone which is your evaluation or reading of an other than the patient, such as a parent, X-ray, ECG, or other study (e.g., “I person- spouse, or group home staff member. (This ally reviewed the X-ray and it shows …”) is included in Category 2 for level 3 data, and can include your personal evaluation but falls into Category 1 for level 4 and 5 of a pertinent study done in the past at data.) your or another facility. It does not include

THE ROSETTA STONE FOUR-STEP TEMPLATE FOR CODING OFFICE VISITS

Level 2 visit Level 3 visit Level 4 visit Level 5 visit Step 1: Total 99212 < 20 min 99213 20-29 min 99214 30-39 min 99215 40-54 min time* 99202 < 30 min 99203 30-44 min 99204 45-59 min 99205 60-74 min Step 2: Level 2 problem: Level 3 problem: two Level 4 problem: one Level 5 problem: one Problems plus minor or self-limited or more minor or self- unstable chronic illness†† chronic illness with problem (e.g., limited problems OR two stable chronic severe exacerbation simple rash) OR one stable illnesses (e.g., diabetes or (e.g., life-threatening PLUS < 20 minutes chronic illness (e.g., chronic kidney disease) COPD/asthma attack) (or < 30 minutes for COPD, diabetes, or OR one undiagnosed new OR one acute or new patients) hypertension) problem with uncertain chronic illness or OR one uncomplicated prognosis (e.g., breast lump) injury that poses acute illness/injury (e.g., a threat to life or OR one acute illness bodily function (e.g., UTI, sinusitis, or simple with systemic symptoms ankle sprain) heart attack, stroke, (e.g., pneumonia, colitis, pulmonary embolism, † PLUS Rx or pyelonephritis, or pelvic acute kidney injury, OTC medication inflammatory disease) or depression with management OR one acute complicated suicidal ideation) injury (e.g., head injury with PLUS admission brief loss of consciousness) decision PLUS Rx Step 3: Simple Level 4 problem data or social PLUS personally interpret determinants a study (e.g., “I personally of health looked at the X-ray and it (SDOH) shows …”) OR discuss patient management or a study with an external physician OR modify workup or treatment due to SDOH Step 4: Level 4 problem Level 5 problem Complex data PLUS review/order unique PLUS order/interpret test/study (one point each), a study (e.g., X-ray or review external notes from ECG) and review/order unique source (one point two tests each), or use independent historian (one point max). At least three points needed.

*Codes 99202-99205 are for new patients and 99212-99215 are for established patients. For visits that take longer than level 5 time, see “Prolonged Services Guide.” †Rx = prescription drug management (ordering, stopping, modifying, or deciding to continue a prescription medication). ††Condition is worsening, poorly controlled, or not to goal (e.g., hypertension with elevated blood pressure or diabetes with elevated A1C). Note: You are not credited with interpretation of a test or study if you are billing separately for it.

24 | FPM | July/August 2021 www.aafp.org/fpm CODING APPROACH

PROLONGED SERVICES GUIDE

Note that Medicare uses higher time thresholds for reporting prolonged services and a different code. Check with your payers to determine which codes and times they are using.

Established patient New patient Prolonged services CPT 99417 x 1 + 99215, 55-69 min 99417 x 1 + 99205 , 75-89 min (99417) 99417 x 2 + 99215, 70-84 min 99417 x 2 + 99205, 90-104 min 99417 x 3 + 99215, > 84 min 99417 x 3 + 99205, > 104 min Prolonged services G2212 x 1 + 99215, 69-83 min G2212 x 1 + 99205, 89-103 min Medicare (G2212) G2212 x 2 + 99215, 84-98 min G2212 x 2 + 99205, 104-118 min G2212 x 3 + 99215, > 98 min G2212 x 3 + 99205, > 118 min

reviewing another clinician’s written which is another level 5 scenario. The report only, and it does not include studies few remaining patient visits that have for which you are also billing separately not already been coded require analyzing for your reading. data (Steps 3 and 4). (See “The Rosetta Stone Category 3 data includes the following: four-step template for coding office visits.”) • Discussion of patient management or Step 1: Total time. Think time first. If test interpretation with an external physi- your total time spent on a visit appropri- cian, other qualified health care profes- ately credits you for level 3, 4, or 5 work, sional, or appropriate source. An external then document that time, code the visit, physician or other qualified health care and be done with it. But if it does not, go to professional is someone who is not in your Step 2. same group practice or specialty. Other Step 2: “Problems plus.” Don’t be afraid appropriate sources could include, for to move on from time-based coding if you example, consulting a patient’s teacher believe you performed a higher level visit about the patient’s attention deficit hyper- using MDM. Many visits can be coded with activity disorder. MDM just by answering these two questions: What was the highest-level problem you A STEP-BY-STEP TIMESAVER The majority of office visits can be opti- mally coded by using time or by looking at what level of problems were addressed The majority of office visits can be (see Steps 1 and 2 below) and whether a pre- optimally coded by using time or by scription medication was involved. A level 3 problem can be coded as a level looking at what level of problems 3 visit if you address it with an OTC or prescription medication. A level 4 problem were addressed. can be coded as a level 4 visit if you order prescription medication or perform any other type of prescription drug manage- addressed during the office visit? And did ment (modifying, stopping, or deciding to you order, stop, modify, or decide to con- continue a medication). Most level 2 and tinue a prescription medication? level 5 office visits are straightforward, and If you addressed a level 2 problem and most level 5 visits will be coded by time. your total time was less than 20 minutes They will typically be visits in which you (or less than 30 for a new patient), then address multiple problems or complicated code level 2. problems and the total time exceeds 40 If you addressed a level 3 problem, plus minutes for established patients. This is you recommended an OTC medication or much more common than seeing critically performed prescription drug management, ill patients who may require admission, then code level 3. ➤ www.aafp.org/fpm July/August 2021 | FPM | 25 If you addressed a level 4 problem, plus Code level 4 if you saw a patient for a level 4 you performed prescription drug manage- problem and did any of the following: ment, then code level 4. • Personally interpret a study (e.g., X-ray), Chronic disease management often • Discuss management or a test with an qualifies as level 4 work. For documenta- external physician, tion, think “P-S-R”: problem addressed, sta- • Modify your workup or treatment tus of the problem (stable vs. unstable), and because of social determinants of health. prescription drug management (Rx). This Step 4: Level 4 or 5 problem with com- trio should make it clear to coders, insur- plex data. If you saw a patient for a level 4 ance companies, and auditors that level 4 problem and still haven’t been able to code work was performed. the visit at this point, you have to tally For instance, if a patient has controlled Category 1 data points: hypertension and diabetes and you docu- • Review/order of each unique test ment that you decided to continue the cur- equals one point each, rent doses of losartan and metformin, that’s • Review of external notes from each level 4 (two stable chronic illnesses plus unique source equals one point each, prescription drug management). If you see • Use of an independent historian equals a patient with even one unstable chronic ill- one point. ness and document prescription drug man- Once you reach three points, code it as agement to address it, that’s also level 4. level 4. For a level 5 problem, if you see a really For a level 5 problem, if you see a really sick patient and decide to admit or consider sick patient, order/interpret an X-ray or admission (and you document your thought ECG, and review/order two lab tests, then process in your note), then code level 5. code level 5. By starting with total time and, if necessary, moving on to “problems plus,” IN SUMMARY you will probably be able to optimally Following these steps should allow you to code 90% of your office E/M visits. But on quickly identify the optimal level to code the rare occasions when you see a patient most any E/M office visit (for pre-op visits, for level 4 or 5 problems for less than the see “Coding pre-ops template.”) required time and don’t do any prescription Here’s a catchy rhyme to remember the drug management, you may have to basic outline of the steps: proceed to Steps 3 and 4. To finish fast, Step 3: Level 4 problem with simple data code by time and problems first, or social determinants of health concerns. and save data for last. By mastering the new coding rules and terminology and applying this four-step CODING PRE-OPS TEMPLATE approach, you can code office visits more quickly, accurately, and confidently — and Level 3 Minor surgery (e.g., insertion of pressure equalizing then spend more time with your patients tubes) without risk factors and less time at the computer.

Level 4 Minor surgery with risk factors* OR major surgery 1. CPT Evaluation and Management (E/M) Office or Other with or without risk factors; PLUS evaluate/review one Outpatient and Prolonged Services Code and Guideline unstable or two stable chronic illnesses OR order/review Changes. American Medical Association. Accessed June three tests/studies (three data points) OR interpret one 10, 2021. https://www.ama-assn.org/system/files/2019- study (e.g., X-ray or ECG) 06/cpt-office-prolonged-svs-code-changes.pdf 2. E/M Office Visit Compendium 2021. American Medical Level 5 Major surgery with risk factors PLUS order/review two Association; 2020. tests and interpret one study (e.g., X-ray or ECG) 3. Table 2 – CPT E/M office revisions level of medical deci- * Risk factors include hypertension, chronic kidney disease, diabetes, coronary artery disease, sion making. American Medical Association. Accessed chronic obstructive pulmonary disease, etc. For documentation, include risk factors in your June 10, 2021. https://www.ama-assn.org/system/ opening statement, e.g., “Mrs. Jones is seen today for [type of surgery] preoperative con- files/2019-06/cpt-revised-mdm-grid.pdf sultation requested by Dr. Smith for patient’s R TKA scheduled on 4/8/21 at Mercy Hospital because of her diabetes and hypertension.” Note: You are not credited with interpretation of a test or study if you are billing separately Send comments to [email protected], or for it. add your comments to the article online.

26 | FPM | July/August 2021 www.aafp.org/fpm MEDICARE PREVENTIVE SERVICES 1. Annual Wellness Visit a. G0402 – IPPE – within the first 12 months of MCR part B coverage i. Once in lifetime. ii. ECG-both parts (G0405). b. G0438 – Initial (once in lifetime). c. G0439 – Subsequent (annual). 2. Advance Care Planning (ACP) a. 99497-ACP including explanation and discussion of advance directives (first 30 minutes-can bill starting at 16 minutes). b. 99498-Each additional 30 minutes. List separately in addition to code for primary procedure. 3. Alcohol misuse screening and counseling a. G0442-Annual alcohol misuse screening, 15 minutes. b. G0443-Brief face-to-face behavioral counseling for alcohol misuse, 15 minutes. Can bill 4 times per year for those who screen positive. 4. DEXA-77080. Every 2 years. 5. Colorectal cancer (CRC) screening. a. 50-85 years of age. b. Asymptomatic. c. Average risk of developing CRC. d. 81528-Cologuard (once every 3 years).

e. Age 50 or older who are at normal risk of developing CRC or high risk of developing CRC. i. Screening colonoscopy (~ every 10 years—no age limitation). 1. G0105 (high risk). 2. G0121 (not high risk). ii. Fecal occult blood tests (FOBT) (G0328 or 82270—once in 12 months). CPT code for FIT = 82274. iii. Flexible sigmoidoscopy (G0104 -- ~ every 4 years). iv. Barium enema (G0106 or G0120 -- ~ every 4 years).

f. High risk—Medicare covers: i. Screening FOBT once in 12 months. ii. Flex sig once in 48 months. iii. Screening colonoscopy once in 24 months. iv. Screening barium enema once in 24 months. 6. Counseling to prevent tobacco use. Two cessation attempts per year. Each attempt may have maximum of 4 intermediate or intensive sessions (total of 8 per year). a. 99406: > 3 minutes; up to 10 minutes (intermediate). b. 99407: > 10 minutes (intensive). 7. Depression screening, annual: G0444 (15 minutes). 8. Diabetes self-management training (DSMT). Provider must be accredited by ADA, AADE, or IHS. Initial year--Up to 10 hours initial training within continuous 12-month period. Subsequent year--Up to 2 hours follow-up training each year after initial year. a. G0108 – Individual. Per 30 minutes. b. G0109 – Group (2 or more). Per 30 minutes. 9. Intensive behavioral therapy (IBT) for obesity. BMI >/= 30. Medicare will pay for up to 22 of these codes in a 12-month period. Face-to-face frequency: Month 1: Weekly. Months 2-6: Every other week. Months 7-12: Every month if cert requirements met (must have lost at least 3 kg). a. G0447 (individual: 15 minutes). b. G0473 (group of 2-10: 30 minutes). 10. Lung cancer screening. a. G0296 (counseling visit-required before initial LDCT). b. G0297 (LDCT). 11. Medical nutrition therapy. 12. Medicare diabetes prevention program (MDPP) expanded model. 13. Prostate cancer screening (annual for age 50+). a. G0102-DRE. b. G0103-PSA. 14. Screening for cervical cancer with HPV tests (G0476). All asymptomatic females 30-65 years old; once every 5 years. 15. Sexually transmitted infection (STI) screening & high-intensity behavioral counseling (HIBC) to prevent STIs (semi-annual; 2 per year; 30 minutes each). Sexually active adolescents and adults at increased risk. a. HIBC = G0445. b. Chlamydia, gonorrhea, syphilis, hepatitis B surface antigen (many codes). 16. Screening mammogram. Females age 35 and older. One baseline mammogram between 35 and 39 years of age. One mammogram per year for age 40+. a. 77063-tomosynthesis. b. 77067-mammogram. 17. Screening Pap tests (many codes and Q0091). Low risk: Every 2 years (23 months have passed). High risk of developing cervical cancer or vaginal cancer or childbearing age with abnormal Pap within the past 3 years: Annually (11 months have passed). 18. Screening pelvic exam-all females (G0101). Cervical or vaginal cancer screening; pelvic and clinical breast exam (CBE). Same frequency as Pap. 19. Ultrasound screening for abdominal aortic aneurysm (AAA). Once in lifetime. 76706.

LABS 1. Lipid panel once every 5 years (80061 – Z13.6). 2. Diabetes screening tests. One screening every 6 months for pre-diabetes, otherwise every 12 months. a. 82947: Quantitative blood glucose except reagent strip. b. 82950: Glucose; post glucose dose. c. 82951: GTT (3 specimens). 3. Hepatitis C virus (G0472). a. High risk behavior, for example, past or current illicit drug use. b. Baby Boomer (born between 1945 and 1965). c. Blood transfusion before 1992. 4. HIV screening (80081, G0432, G0433, G0435, or G0475). a. Anyone who asks for the test. i. Annually for ages 15 to 65 years of age without regard to risk. ii. Annually outside these ages if increased risk. b. Pregnant females (3 times during the pregnancy). 5. PSA – See #13.

VACCINATIONS 1. Hepatitis B. Only for those at high (includes patients with diabetes) or intermediate (includes healthcare workers) risk. a. Vaccine (90746). b. Admin (G0010). 2. Influenza. a. Vaccine (many codes). b. Admin (G0008). 3. Pneumococcal pneumonia (PPV). Initial then different pneumococcal vaccine 1 year later. a. PCV-13 vaccine (90670). b. PPSV-23 vaccine (90732). c. Admin (G0009).

Per Novitas website—March, 2021.

Medication Affordability

Objectives: Provide resources to assist with patient access to necessary medications

• Medicare beneficiaries (Part D overview)

• Commercial insurance beneficiaries

• Medicaid or MSCAN beneficiaries

• Self-pay patients

Resources: https://q1medicare.com/PartD-SearchPDPMedicare-2021PlanFinder.php#results

MMIT Coverage Search Application medicaid.ms.gov/providers/pharmacy/preferred-drug-list/

rxassist.org

rxoutreach.org

$4 Prescriptions - Walmart.com

Journal of the American College of Cardiology Vol. 63, No. 2, 2014 Ó 2014 by the American College of Cardiology Foundation ISSN 0735-1097/$36.00 Published by Elsevier Inc.

CORRESPONDENCE

Research Correspondence Prevalence and Outcomes of Unoperated Patients With Severe Symptomatic Mitral Regurgitation and Heart Failure Comprehensive Analysis to Determine the Potential Role of MitraClip for This Unmet Need

To the Editor: Mitral valve (MV) surgery is recommended in unoperated patients with symptomatic severe FMR with a good- patients with severe symptomatic mitral regurgitation (MR) (1). quality echocardiogram available for review, 171 patients (36%) The role of MV surgery is unclear in patients with severe MR would have been eligible for MitraClip based on published criteria. secondary to left ventricular (LV) dysfunction (1). Many patients Over a mean follow-up period of 4.5 years, an increase in use of with severe MR are at high surgical risk due to advanced age, LV the following medical therapies was observed in unoperated patients dysfunction, or comorbidities. The MitraClip device (Abbott with MR: angiotensin-converting enzyme inhibitors (31% to 63%), Vascular, Abbott Park, Illinois) provides a novel percutaneous angiotensin receptor blockers (13% to 25%), beta-blockers (35% option for patients with severe MR with significant improvement to 82%), and aldosterone antagonists (13% to 33%). Cardiac (2,3). However, the characteristics and outcomes of this patient population remain unclear. We sought to assess the characteristics and outcomes of patients with severe symptomatic MR who do not undergo MV surgery and their potential eligibility for MitraClip. Our institutional databases showed that 5,737 patients with se- vere (3þ) MR were identified between January 1, 2000, and December 31, 2008. A search of the electronic medical records yielded 1,095 patients with a diagnosis of severe MR and heart failure. Patients undergoing MV surgery until September 30, 2011, were identified. Exclusion criteria included MV area 2cm2, aortic regurgitation of grade 2/4 or greater, aortic peak velocity 2.5 m/s, hypertrophic cardiomyopathy, active endo- carditis, concomitant aortic valve, aortic or pericardial surgeries, left ventricular assist device implantation, or heart transplantation. Diagnosis and procedures were confirmed by chart review. Patients with unoperated functional MR (FMR) were evaluated regarding eligibility for MitraClip based on published criteria. FMR (814 of 1,095 [74%]) and degenerative MR (DMR) (226 of 1,095 [21%]) were the predominant mechanisms of MR. MV surgery was performed in 47% (518 of 1,095) of patients, whereas 53% (577 of 1,095) of patients were unoperated and medically managed (Fig. 1A). The etiology was FMR and DMR in 57% (294 of 518) and 37% (190 of 518) of the operated and 90% (520 of 577) and 6% (36 of 577) of the unoperated patients, respectively. In the FMR group, 36% (294 of 814) of patients underwent MV surgery and 64% (520 of 814) were medically managed, whereas 84% (190 of 226) of patients with DMR underwent MV surgery and 16% (36 of 226) were medically managed. Unoperated patients had a lower ejection fraction (27 15% vs. 42 17%, p < 0.0001 [Student t test]) and a higher Prevalence and Outcomes of Unoperated Patients With Figure 1 Society of Thoracic Surgeons score compared with operated Severe Symptomatic MR and Heart Failure patients (median [interquartile range]: 5.8 [2.8 to 11.5] vs. 4.0 [2.1 to 7.4], p < 0.001 [Student t test]). (A) Pie chart showing mechanism and management of 1,095 patients with severe The overall 1-year and 5-year mortality rate in unoperated patients symptomatic MR. (B) Mortality and rates of hospitalization for heart failure in unoperated patients with severe MR. DMR ¼ degenerative mitral regurgitation; was 20% and 50%, respectively (Fig. 1B). In these unoperated pa- FMR ¼ functional mitral regurgitation; MR ¼ mitral regurgitation; MV ¼ mitral tients, the proportion of surviving patients hospitalized for heart valve. Figure by Craig Skaggs. failure increased from 41% in the first year to 90% by 5 years. Of 474 186 Correspondence JACC Vol. 63, No. 2, 2014 January 21, 2014:185–90 resynchronization therapy was instituted in 17% (100 of 577) and Nicholas G. Smedira, MDk percutaneous coronary revascularization in 10% (58 of 577) of Guilherme H. Oliveira, MDy unoperated patients. At the end of follow-up, the severity of MR Benico Barzilai, MDy improved to less than grade 2þ in 15% of patients (89 of 577). Venu Menon, MDy In the Euro Heart Survey (396 patients with severe MR), 56% *Samir R. Kapadia, MDy and 32% of patients with DMR and FMR, respectively, did not *Department of Cardiovascular Medicine undergo surgery (4). In contrast, we found that the majority of Cleveland Clinic Foundation patients with DMR underwent surgery and the vast majority (90%) J2-3, 9500 Euclid Avenue of unoperated patients had FMR. This likely reflects the low risk Cleveland, Ohio 44195 and better durability of surgical MV repair for DMR compared E-mail: [email protected] with FMR in a center with surgical expertise and excellence. In contrast, many patients with FMR do not undergo MV surgery http://dx.doi.org/10.1016/j.jacc.2013.08.723 unless coronary artery bypass grafting is feasible and indicated (77% From the yDepartment of Cardiovascular Medicine, Cleveland in our cohort of patients with ischemic FMR) when concomitant Clinic Foundation, Cleveland, Ohio; zDepartment of Internal MV surgery is usually performed. Medicine, Cleveland Clinic Foundation, Cleveland, Ohio; xCase The current class IIb recommendation for MV surgery in the Western Reserve University School of Medicine, Cleveland, Ohio; presence of LV dysfunction is based on lack of survival benefit and a and the kDepartment of Cardiothoracic Surgery, Cleveland Clinic high recurrence rate of MR (5). The outcomes of medically managed Foundation, Cleveland, Ohio. patients, however, remain understudied. Our observations highlight Please note: The Cleveland Clinic receives funding for clinical research from Abbott the poor outcomes of medically managed patients with severe FMR Vascular. Dr. Gillinov is a consultant for Abbott, Edwards, Medtronic, On-X, and with a 1-year mortality rate of 20%, 5-year mortality rate of 50%, AtriCure; and has received honoraria from Intuitive Surgical. All other authors have and high rate of heart failure hospitalization. Until recently, medi- reported that they have no relationships relevant to the content of this paper to disclose. cally managed patients with FMR had no good options short of advanced mechanical support and heart transplantation. MitraClip provides an important new tool for FMR with good clinical REFERENCES improvement and safety (30-day mortality rate of 1%) (2). Non- fi 1. Bonow RO, Carabello BA, Chatterjee K, et al. 2008 focused update randomized studies have shown a clinical bene t of MitraClip in incorporated into the ACC/AHA 2006 guidelines for the manage- high-risk patients with FMR (3). Our study identifies a potentially ment of patients with valvular heart disease. J Am Coll Cardiol 2008; large unmet need for transcatheter MR reduction strategies in 52:e1–142. medically managed patients. In this context, the ongoing studies 2. Feldman T, Foster E, Glower DD, et al. Percutaneous repair or surgery for mitral regurgitation. N Engl J Med 2011;364:1395–406. that are enrolling high surgical risk patients with severe FMR will 3. Whitlow PL, Feldman T, Pedersen WR, et al. Acute and 12-month provide data on the effectiveness of MitraClip in these patients. results with catheter-based mitral valve leaflet repair: the EVEREST II This was a single-center observational study from a large tertiary (Endovascular Valve Edge-to-Edge Repair) High Risk Study. J Am Coll – referral center. It is possible that some patients with severe MR Cardiol 2012;59:130 9. 4. Mirabel M, Iung B, Baron G, et al. What are the characteristics of were hospitalized outside our network of hospitals, resulting in an patients with severe, symptomatic, mitral regurgitation who are denied underestimation of hospitalization rates. Not all patients were surgery? Eur Heart J 2007;28:1358–65. necessarily referred to a cardiothoracic surgeon. However, this 5. Mihaljevic T, Lam BK, Rajeswaran J, et al. Impact of mitral valve mirrors current clinical practice, where the role of surgery for FMR annuloplasty combined with revascularization in patients with functional – remains nebulous. Finally, anatomic criteria for MitraClip are ischemic mitral regurgitation. J Am Coll Cardiol 2007;49:2191 201. evolving with increasing experience, underestimating the potential applicability of this therapy. In conclusion, our analysis not only highlights a substantial unmet need for MitraClip for symptomatic patients with severe Letters to the Editor MR but also provides some benchmark data on their outcomes with medical management.

Sachin S. Goel, MDy Positron Emission Tomography/ Navkaranbir Bajaj, MDz Computed Tomography for Bhuvnesh Aggarwal, MDz Supriya Gupta, MDy Diagnosis of Prosthetic Kanhaiya Lal Poddar, MDy Mobolaji Ige, MDz Valve Endocarditis Hazem Bdair, MDy 18 Abed Anabtawi, MDy Increased Valvular F-Fluorodeoxyglucose Shiraz Rahim, BSx Patrick L. Whitlow, MDy Uptake as a Novel Major Criterion E. Murat Tuzcu, MDy Brian P. Griffin, MDy William J. Stewart, MDy We read with great interest the recent publication by Saby et al. (1), Marc Gillinov, MDk in which 72 patients with suspected prosthetic heart valve (PHV) Eugene H. Blackstone, MDk endocarditis were prospectively studied to determine the accuracy Articles

Percutaneous renal denervation in patients with treatment-resistant hypertension: fi nal 3-year report of the Symplicity HTN-1 study

Henry Krum, Markus P Schlaich, Paul A Sobotka, Michael Böhm, Felix Mahfoud, Krishna Rocha-Singh, Richard Katholi, Murray D Esler

Summary Lancet 2014; 383: 622–29 Background Renal denervation (RDN) with radiofrequency ablation substantially reduces blood pressure in patients Published Online with treatment-resistant hypertension. We assessed the long-term antihypertensive eff ects and safety. November 7, 2013 http://dx.doi.org/10.1016/ Methods Symplicity HTN-1 is an open-label study that enrolled 153 patients, of whom 111 consented to follow-up for S0140-6736(13)62192-3 36 months. Eligible patients had a systolic blood pressure of at least 160 mm Hg and were taking at least three This online publication has been corrected. The corrected version antihypertensive drugs, including a diuretic, at the optimum doses. Changes in offi ce systolic blood pressure and fi rst appeared at thelancet.com safety were assessed every 6 months and reported every 12 months. This study is registered with ClinicalTrials.gov, on February 14, 2014 numbers NCT00483808, NCT00664638, and NCT00753285. See Comment page 583 See Perspectives page 591 Findings 88 patients had complete data at 36 months. At baseline the mean age was 57 (SD 11) years, 37 (42%) patients Monash Centre of were women, 25 (28%) had type 2 diabetes mellitus, the mean estimated glomerular fi ltration rate was Cardiovascular Research and 85 (SD 19) mL/min per 1·73 m², and mean blood pressure was 175/98 (SD 16/14) mm Hg. At 36 months signifi - Education in Therapeutics, cant changes were seen in systolic (–32·0 mm Hg, 95% CI –35·7 to −28·2) and diastolic blood pressure (–14·4 mm Hg, School of Public Health and Preventive Medicine, Monash –16·9 to –11·9). Drops of 10 mm Hg or more in systolic blood pressure were seen in 69% of patients at 1 month, University, Alfred Hospital, 81% at 6 months, 85% at 12 months, 83% at 24 months, and 93% at 36 months. One new renal artery stenosis Melbourne, VIC, Australia requiring stenting and three deaths unrelated to RDN occurred during follow-up. (Prof H Krum PhD); Neurovascular Hypertension and Kidney Disease Laboratory, Interpretation Changes in blood pressure after RDN persist long term in patients with treatment-resistant Baker IDI Heart and Diabetes hypertension, with good safety. Institute and Heart Centre, Alfred Hospital, Melbourne, Funding Ardian LLC/Medtronic Inc. VIC, Australia (Prof M P Schlaich MD); Davis Heart and Lung Research Introduction research into the role of adrenergic overdrive in hyper- Institute, Ohio State Systemic hypertension is the single largest contributor to tension described increased norepinephrine concen- University, Columbus, OH, USA death worldwide.1 One in three adults (around 1 billion trations in plasma that arose from renal and systemic (Prof P A Sobotka MD); 9,10 Kardiologie, Angiologie und people) worldwide are aff ected and the number is spillover into the circulation. Internistische Intensivmedizin, expected to increase to 1·6 billion by 2025.2,3 Hypertension Renal denervation (RND) by radiofrequency ablation Universitätsklinium des strikingly increases the risk of stroke, myocardial infarc- has been associated with reductions in blood pressure.11–13 Saarlandes, Homburg, 3,4 Germany (M Böhm MD, tion, heart failure, and kidney disease. A patient with treatment-resistant hypertension who F Mahfoud MD); Prairie Heart Despite the use of multiple antihypertensive drugs, presented with notable norepinephrine spillover had Institute at St John’s Hospital, including diuretics, at recommended or target doses, substantially reduced whole-body norepinephrine con- Springfi eld, IL, USA hypertension remains uncontrolled in a substantial pro- centrations and reduced blood pressure after RDN.14 A (K Rocha-Singh MD, 5,6 R Katholi MD); and Baker IDI portion of patients. Multiple factors can contribute to substantial reduction in central sympathetic outfl ow has 15 Heart and Diabetes Institute, poor control of blood pressure: suboptimum pharma- also been reported. RDN, therefore, off ers a therapeutic Melbourne, VIC, Australia ceutical care and patients deciding not to commit to option for the management of treatment-resistant hyper- (Prof M D Esler MBBS) lifelong polypharmacy. Furthermore, drugs are associated tension for patients who cannot attain targeted blood Correspondence to: with adverse clinical events and patients frequently do pressures with medications alone.16 Prof Henry Krum, Centre of not adhere to therapy because of side-eff ects, fi nancial In a proof-of-concept study of RDN (Symplicity Cardiovascular Research and 7 17 Education in Therapeutics, concerns, or a lack of hypertensive symptoms. Finally, HTN-1 ), we found signifi cant substantial reductions in Department of Epidemiology antihypertensive medications do not lower blood pres- blood pressure by 1 month after treatment that continued and Preventive Medicine, School sure in some patients despite full adherence and attentive to the 12 month endpoint. The longer-term durability of of Public Health and Preventive Medicine, Monash University/ health care. the treatment response has been questioned because, Alfred Hospital, Melbourne, Strategies that target the contribution of overactivity of theoretically, the treated renal nerves could regrow and VIC 3004, Australia the sympathetic nervous system have proven clinically regain function or a counter-regulatory response might [email protected] important in several disorders.8 Patients with hype- develop.18–20 For this reason, we extended follow-up to rtension might benefi t from agents that inhibit central 36 months specifi cally to assess the durability of blood- release of catecholamines, or from use of β and α pressure-lowering eff ects and investigate any late adverse blockers, which inhibit catecholamine receptors. Early vascular or renal eff ects.

622 www.thelancet.com Vol 383 February 15, 2014 Downloaded for Anonymous User (n/a) at West Virginia School of Osteopathic Medicine from ClinicalKey.com by Elsevier on August 17, 2021. For personal use only. No other uses without permission. Copyright ©2021. Elsevier Inc. All rights reserved. Articles

Methods 2 min each were applied. After each delivery the catheter Patients was drawn back by at least 5 mm and circumferentially Symplicity HTN-117 is an open-label cohort study that rotated to ensure disruption of the sympathetic plexus enrolled 153 patients into four protocols at 19 centres in surrounding the renal artery. Australia, Europe, and the USA. Detailed study methods Blood pressure was measured during follow-up offi ce have been reported previously.17 Briefl y, patients were visits. Adverse events, blood chemistry (including renal required to have an offi ce systolic blood pressure of function), and vital signs were assessed per protocol at 160 mm Hg or higher despite treatment with at least each visit. Patients underwent renal angiography before three antihypertensive drugs, including one diuretic, RDN, and renal imaging (duplex scan or angiography) or confi rmed intolerance to medications. Estimated was done at 6 months and at various timepoints after the glomerular fi ltration rate (eGFR), based on creatinine procedure to assess the renal arteries for pathological concentrations in serum, was required to be at least changes. From 12 months onwards we assessed changes 45 mL/min per 1·73 m². We took care not to enrol in sitting offi ce systolic and diastolic blood pressures patients with renovascular abnormalities, including every 12 months. renal-artery stenosis, previous renal stent or angioplasty, dual renal arteries, or polar arteries. In the fi rst 12 months Statistical analysis after RDN investigators were encouraged not to change Mean (95% CI) changes from baseline in blood pressure background antihypertensive medicine unless clinically and eGFR at 1, 3, 6, 12, 24, and 36 months were calculated. indicated. After 12 months, patients chose whether to Changes in offi ce blood pressure from baseline to each give consent to be followed up for 24 or 36 months, follow up timepoint were assessed by the paired t test. dependent on the protocol to which they were assigned We deemed p values of 0·05 or lower to be signifi cant. (fi gure 1). During extended follow-up medications could All statistical analyses were done with SAS (version 9.2). be changed as required. 111 patients gave consent to be We set 10 mm Hg as the threshold for clinically relevant followed up for 36 months. response because the time to response is of interest. We

Study procedure The percutaneous RDN procedure has been described Enrolled Patients previously.17 Briefl y, a Symplicity renal denervation patients followed up (n=150)* to 36 months catheter (Medtronic, Santa Rosa, CA, USA) was intro- (n=88) duced into each renal artery via the femoral artery. Demographics Multiple radiofrequency ablations of 8 W or less for up to Age (years) 57·1 (11·1) 57·0 (11·4) Female sex (%) 56 (38%) 37 (42%) Non-white ethnic origin (%) 7 (5%) 4 (5%) 153 patients enrolled Comorbidities 3 missing baseline blood-pressure data* Type 2 diabetes mellitus (%) 47 (31%) 25 (28%) Coronary artery disease (%) 33 (22%) 20 (23%) 150 patients eligible Hyperlipidaemia (%) 92 (61%) 62 (71%) 18 excluded 2 died Estimated GFR (mL/min per 1·73m²) 83·4 (19·7) 84·6 (18·9) 4 withdrew Blood pressure 7 lost to follow-up 5 missed visit or incomplete data Baseline systolic (mm Hg) 175·1 (15·9) 174·6 (12·5) 132 (96%) of 137† Number of antihypertensive medications 5·0 (1·7) 5·2 (1·7) assessed at 12 months Diuretic (%) 138 (92%) 82 (93%) 32 excluded 23 completed study at 12 months Aldosterone antagonist (%) 41 (27%) 25 (28%) 1 died Angiotensin-receptor blocker (%) 102 (68%) 61 (69%) 1 withdrew ACE inhibitor (%) 80 (53%) 45 (51%) 1 lost to follow-up 6 missed visit or incomplete data Direct renin inhibitor (%) 25 (17%) 16 (18%) 105 (95%) of 111† β blocker (%) 126 (84%) 76 (86%) assessed at 24 months Calcium-channel blocker (%) 120 (80%) 70 (80%) 23 excluded 12 completed study at 24 months Centrally acting sympatholytic (%) 54 (36%) 33 (38%) 1 withdrew Vasodilator (%) 32 (21%) 16 (18%) 4 lost to follow-up 6 missed visit or incomplete data α-1 adrenergic blocker (%) 32 (21%) 19 (22%) 88 (94%) of 94† assessed at 36 months Data are mean (SD) or number (%). GFR=glomerular fi ltration rate. ACE=angiotensin-converting enzyme. *Baseline blood pressure values missing for three patients. Figure 1: Trial profi le *Included in safety analysis. †Numerators include patients who consented to be Table 1: Characteristics of patients at baseline and 36 months followed and who completed follow-up to the timepoint stated.

www.thelancet.com Vol 383 February 15, 2014 623 Downloaded for Anonymous User (n/a) at West Virginia School of Osteopathic Medicine from ClinicalKey.com by Elsevier on August 17, 2021. For personal use only. No other uses without permission. Copyright ©2021. Elsevier Inc. All rights reserved. Articles

Systolic blood pressure Diastolic blood pressure 0

Mean Mean (SD) Mean Mean (SD) –5 (SD) change from (SD) change from –10 –11 –12 pressure baseline (95% CI) pressure baseline (95% CI) –13 –10 –14 (mm Hg) (mm Hg) –15 Baseline 175·1 ·· 97·7 ·· –21 (n=150) (15·9) (14·3) –20 1 month 156·1 −18·9 (19·2) 88·2 −9·4 (12·1) –26 –27 (n=141) (20·9) (−22·1 to –15·7) (13·7) (−11·4 to –7·4) –25 –30

6 months 152·7 −22·0 (21·9) 87·3 −10·2 (13·6) BP change (mm Hg) –32 (n=144) (21·1) (−25·7 to –18·4) (13·2) (−12·4 to –7·9) –30 12 months 149·1 −26·5 (21·5) 84·3 −13·5 (14·1) Systolic BP (n=132) (22·6) (−30·2 to –22·8) (13·5) (−15·9 to –11·1) –35 Diastolic BP 24 months 145·4 −28·9 (23·5) 82·9 −14·0 (15·2) p<0·01 for all timepoints (n=105) (21·7) (−33·5 to –24·4) (13·0) (−16·9 to –11·1) –40 1 (n=80) 6 (n=88) 12 (n=85) 24 (n=82) 36 (n=88) 36 months 142·6 −32·0 (17·6) 82·0 −14·4 (11·8) Time (months) (n=88) (17·7) (−35·7 to −28·2) (11·0) (−16·9 to −11·9)

Table 2: Changes in blood pressure from baseline at follow-up visits Figure 2: Change from baseline in offi ce blood pressure in patients who completed 36 months of follow-up Data are mean (error bars show 95% CI). BP=blood pressure. used reductions of 20 mm Hg or more as a more rigor- ous measure. We also compared response to RDN in ≥180 mm Hg 140–159 mm Hg subgroups of patient, by age (older than 65 years vs 160–179 mm Hg <140 mm Hg 65 years and younger), renal function (eGFR higher than 100 60 vs 45–60 mL/min per 1·73 m²), and diabetes status 90 (type 2 diabetes vs no diabetes). 80 70 Role of funding source 60 The sponsor designed the study in collaboration with the 50 study investigators and was responsible for data collec- 40 tion and data analysis. The authors are responsible for 30 Proportion of patients (%) Proportion data interpretation and writing of the report. The 20 corresponding author had full access to all the study data 10 and had fi nal responsibility for the decision to submit. 0 Baseline 1 12 24 36 Results (n=150) (n=141) (n=132) (n=105) (n=88) Time (months) 153 patients with treatment-resistant hypertension were enrolled in the Symplicity HTN-1 study, of whom 88 had Figure 3: Distribution of changes in systolic blood pressure for all complete data at 36 months (fi gure 1). Patients did not treated patients diff er signifi cantly at baseline and at 36 months for

demographic characteristics, comorbidities, blood pres- 100 SBP drop ≥10 mm Hg SBP drop ≥20 mm Hg sure, number of antihypertensive medications, and use 93% 90 85% 81% 83% of antihypertensive drugs (table 1). 77% 80 74% Patients received an average of 4·0 ablations per artery 69% 70 66% 63% (range 1·0–6·0). Intravenous narcotics and sedatives 59% were used to manage pain during the delivery of 60 radiofrequency energy. The mean total procedure time 50 from initial femoral access to withdrawal of the catheter 40 was 66·1 (SD 23·0) min. 30 There were no catheter or generator malfunctions 20 Proportion showing response (%) and no major clinical complications associated with 10 RDN. Complications were reported in four (2·6%) of 0 153 patients: one renal-artery dissection occurred during 1 (n=80) 6 (n=88) 12 (n=85) 24 (n=82) 36 (n=88) Time (months) catheter delivery before the application of radiofrequency energy, and three patients had access-related compli- Figure 4: Proportions of patients assessed to 36 months who showed cations in the groin. All events were treated without treatment responses at diff erent timepoints in the study Treatment response was defi ned as reductions of 10 mm Hg or more in systolic further sequelae. Eight (5·2%) patients had episodes of blood pressure, and reductions of 20 mm Hg or more were used as a more bradycardia associated with the ablation procedure, but rigorous measure of response.

624 www.thelancet.com Vol 383 February 15, 2014 Downloaded for Anonymous User (n/a) at West Virginia School of Osteopathic Medicine from ClinicalKey.com by Elsevier on August 17, 2021. For personal use only. No other uses without permission. Copyright ©2021. Elsevier Inc. All rights reserved. Articles

no episodes occurred after catheter removal, and no A episodes of vasovagal syncope were reported. No patient 0 experienced orthostatic hypotension after RDN. –10 –9 Possible renal-artery stenosis was investigated in four –5 –10 (2·6%) patients. Two have been previously reported and –10 –15 –16 included one progression of a pre-existing stenosis to 50% –16 –15 –25 unrelated to RDN that was stented without further –26 sequelae, and one was a moderate stenosis that was not –20 –27 –31 haemodynamically relevant and did not require treatment. –30 –25 One patient had a new 70–80% stenosis of the left renal –33 artery seen on duplex ultrasonography at 18 months, but it –30 was found to be only 20–30% on follow-up angiography. –35

The fourth case occurred in a patient with a history of Office BP change (mm Hg) peripheral artery disease, coronary artery disease, lymph- –40 oma, hypercholesterolaemia, and renal insuffi ciency who –45 had presented with recurrent high blood pressure and developed an angiographically confi rmed new 80% sten- –50 Systolic BP Diastolic BP osis of the right renal artery 24 months after RDN. The –55 stenosis was successfully stented. <65 years ≥65 years <65 years ≥65 years <65 years ≥65 years Three patients died during follow-up, but all deaths (n=98) (n=34) (n=78) (n=27) (n=65) (n=23) were unrelated to the device or treatment: one myocardial 12 months 24 months 36 months infarction on day 3, one sudden cardiac death after the B 6-month visit, and one cardiac and respiratory arrest after 0 the 18-month visit. At 18 months one patient had a hypotensive event with acute renal failure related to –5 severe left-leg cellulitis that required admission to –13 –22 –14 –10 –19 –15 hospital for antibiotics and intravenous fl uids. The –33 –15 –23 episode resolved and was deemed unrelated to RDN. –28 Another patient was admitted to hospital with a hypo- –20 –37 –27 tensive event associated with severe diarrhoea and –29 –25 dehydration. The episode resolved after treatment, with –32 no sequelae and was also not attributed to RDN. One –30 patient experienced orthostatic hypotension on two occasions (31 and 32 months after RDN) that were –35 Office BP change (mm Hg) managed by discontinuation of an antihypertensive drug. –40 13 patients were admitted to hospital for hypertensive –45 episodes, but all were treated and discharged without complications. –50 One patient presented at 24 months with acute renal –55 failure and dehydration secondary to vomiting and eGFR 45–60 eGFR >60 eGFR 45–60 eGFR >60 eGFR 45–60 eGFR >60 diarrhoea after undergoing surgery for breast recon- (n=10) (n=118) (n=9) (n=92) (n=6) (n=78) struction. She was diagnosed as having acute tubular 12 months 24 months 36 months necrosis that was attributed to multiple antihypertensive Follow-up agents, including hydrochlorothiazide, spironolactone, Figure 5: Change in offi ce blood pressure by subgroup and aliskiren. Although she temporarily required haemo- (A) Age and (B) renal function, expressed as eGFR (mL/min per 1·73 m²). Data are mean (95% CI). BP=blood dialysis at the time of the event, renal function recovered pressure. eGFR=estimated glomerular fi ltration rate. to normal (creatinine concentration in serum 71 μmol/L and blood urea nitrogen 6·1 mmol/L). RDN, 55 (69%) of 80 patients had reductions in systolic Reductions in systolic and diastolic blood pressure that blood pressure of at least 10 mm Hg, which rose pro- were seen by 12 months persisted to 36 months (table 2, gressively to 82 (93%) of 88 at 36 months. Reductions of fi gure 2). Heart rate did not change signifi cantly from 20 mm Hg or more were seen in 68 (77%) of these baseline to 36 months. The proportions of patients with 88 patients (fi gure 4). systolic blood pressure of 180 mm Hg or higher decreased Blood pressure responses at 12, 24, and 36 months did over the duration of the study, from 30% at baseline to 5% not diff er signifi cantly between age groups or by renal at 36 months. The proportion who achieved target systolic function (fi gure 5). Similarly, changes in blood pressure blood pressure values of less than 140 mm Hg increased from baseline for patients with and without type 2 signifi cantly at all timepoints (fi gure 3). At 1 month after diabetes were similar (data not shown). www.thelancet.com Vol 383 February 15, 2014 625 Downloaded for Anonymous User (n/a) at West Virginia School of Osteopathic Medicine from ClinicalKey.com by Elsevier on August 17, 2021. For personal use only. No other uses without permission. Copyright ©2021. Elsevier Inc. All rights reserved. Articles

Baseline (n=88) 6 months (n=86) 12 months (n=85) 24 months (n=80) 36 months (n=88) Diuretic 82 (93%) 81 (94%) 80 (94%) 75 (94%) 84 (95%) Aldosterone antagonist 25 (28%) 26 (30%) 26 (31%) 26 (33%) 29 (33%) Angiotensin-receptor blocker 51 (69%) 59 (69%) 60 (71%) 58 (73%) 63 (72%) ACE inhibitor 45 (51%) 46 (53%) 45 (53%) 43 (54%) 49 (56%) Direct renin inhibitor 16 (18%) 15 (17%) 15 (18%) 14 (18%) 17 (19%) β blocker 76 (86%) 74 (86%) 75 (88%) 72 (90%) 80 (91%) Calcium-channel blocker 70 (80%) 71 (83%) 69 (81%) 65 (81%) 74 (84%) Centrally acting sympatholytic 33 (38%) 33 (38%) 33 (39%) 32 (40%) 35 (40%) Vasodilator 16 (18%) 16 (19%) 16 (19%) 15 (19%) 17 (19%) α-1 adrenergic blocker 19 (22%) 18 (21%) 19 (22%) 19 (24%) 21 (24%)

ACE=angiotensin-converting enzyme.

Table 3: Medication use, by class, during the study

The average number of antihypertensive medications These long-term fi ndings are of mechanistic interest and used at baseline was 5·1 (SD 1·7) for the overall are clinically important. First, they indicate no functional population and 5·2 (1·7) for the patients assessed at re-innervation or any counter-regulatory mechanisms 36 months. At 6 months after RDN the mean number of develop over time that could lessen the effi cacy of the medications taken was 5·1 (1·6), at 24 months was procedure. Additionally, although blood pressure did not 5·4 (1·7), and at 36 months was 5·6 (1·6). The proportions drop immediately after RDN in all patients, the response of medication use, by class, at 36 months were similar to rate rose with follow-up longer than 12 months. The those at baseline (table 3). response defi nition we selected for reduction in blood Concentrations of sodium, potassium, chloride, and pressure (10 mm Hg or more systolic) is somewhat glucose in serum were within normal limits at base- arbitrary, but we believe it represents a clinically relevant line and remained essentially unchanged throughout drop in blood pressure in patients with treatment-resistant follow-up. Mean creatinine concentrations in serum hypertension. The potential benefi t of RDN is further increased progressively from 83·8 (SD 20·1) μmol/L to indicated by our fi nding that 68 (77%) of 88 patients had 92·0 (32·5) μmol/L over the 36 months (p=0·05). This reductions of at least 20 mm Hg in systolic blood pres- change was associated with a corresponding decrease in sure at 36 months. A meta-analysis by Lewington and mean eGFR, from 83·6 (SD 19·7) to 74·3 (28·0) mL/min colleagues21 showed a doubling of cardiovascular mortality per 1·73 m² (p=0·05). 28 patients experienced a fall in risk for each 20/10 mm Hg increase in blood pressure, eGFR of more than 25% at one or more timepoints after which suggests that the reductions in our study would RDN. These decreases were transient in 16 (57%) confer a meaningful benefi t in patients whose blood patients, with renal function having improved at the next pressures remain above the target range despite already follow-up visit. Of the 12 (43%) patients with decreased having been substantially lowered. Furthermore, half of eGFR on more than one occasion, baseline eGFR values patients assessed at 36 months had blood pressures lower were higher than 100 mL/min per 1·73 m² in fi ve, and in than 140 mm Hg. The reasons for delayed blood-pressure four of these fi ve eGFR values never fell below responses in some patients after RDN are unclear. We 60 mL/min per 1·73 m² (eGFR was 57 mL/min per 1·73 m² speculate that the timecourse of vascular remodelling, at one visit for the fi fth patient). Three patients had eGFR delay in resetting of the barorefl ex, or changes in renin- lower than 45 mL/min per 1·73 m². angiotensin-system activity might contribute. Further research will be needed to address this question. Discussion Whether or not specifi c subgroups of patients are Percutaneous RDN with radiofrequency ablation is safe, particularly responsive to RDN is an important con- eff ective, and leads to persistent reductions in blood sideration. The numbers in this study alone are too small pressure in patients with severe treatment-resistant to undertake a meaningful analysis of potential pre dictors. hypertension. In most (93%) patients assessed at In another study, however, of ambulatory and offi ce blood 36 months, initial blood-pressure responses were main- pressures after RDN in patients with treatment-resistant tained, although the potential role of modifi cations to hypertension, offi ce systolic blood pressure at baseline antihypertensive drug therapy is unclear. Advanced age, was the only independent correlate of response.22 We renal impairment, and diabetes status had no substantial assessed the eff ects of age, renal function, and diabetes eff ects on treatment response or time to treatment status because some mechanistic diff erences between eff ect. We saw no evidence of major vascular or renal subgroups could plausibly aff ect blood pressure. We noted adverse events. no signifi cant diff erences between any subgroups and,

626 www.thelancet.com Vol 383 February 15, 2014 Downloaded for Anonymous User (n/a) at West Virginia School of Osteopathic Medicine from ClinicalKey.com by Elsevier on August 17, 2021. For personal use only. No other uses without permission. Copyright ©2021. Elsevier Inc. All rights reserved. Articles

therefore, RDN might be benefi cial in patients with treatment-resistant hypertension irrespective of these Panel: Research in context features. Nevertheless, effi cacy will need to be confi rmed Systematic review in these and other subgroups in larger studies or through Overactivity of the sympathetic nervous system plays an important part in the assessment of data collected in planned registry activities. development of hypertension. Studies in normotensive individuals and hypertensive Concerns have been expressed about potential late patients have shown increased rates of norepinephrine spillover from renal sympathetic vascular complications and biochemical disruption after nerves to plasma and increased rates of sympathetic nerve fi ring in hypertensive patients RDN. The only vascular events were expected peri- compared with normotensive individuals.34 Since 1948, surgical sympathectomy, such as procedural local events related to femoral-artery haemo- thoracolumbar sympathectomy, splanchnicectomy, and coeliac ganglionectomy, has led stasis. These were all managed according to current to substantial reductions in blood pressure, which provides clear evidence that the medical practice without complications. Importantly, the sympathetic nervous system is an important contributor to the pathogenesis of high rate of stenosis in late follow-up was very low. The blood pressure.35,36 Adverse eff ects, however, ultimately led to treatment with incidence was consistent with rates in severely hyper- pharmacological options. Blood-pressure reductions have been reported after bilateral tensive populations and, although not thought to be nephrectomy in patients with end-stage renal disease and in animal studies of renal related to RDN, cannot be entirely dissociated from the sympathectomy.37,38 Radiofrequency ablation was developed to provide a minimally procedure.23 We found no evidence of other vascular invasive, safe, and eff ective alternative treatment for severely hypertensive patients who pathologies, which is consistent with animal studies in are unresponsive to traditional pharmacological therapies. We searched PubMed for which minimum endothelial disruption, rapid re- studies published between January, 2009, and January, 2013, on the use of this procedure endothelialisation in the renal artery, and no evidence of with the terms “percutaneous renal denervation”, “sympathetic”, “clinical trial”, and late stenosis were seen after RDN.24 “resistant hypertension”. We identifi ed 14 clinical trials and one randomised trial of Although eGFR was well preserved overall, renal radiofrequency RDN for resistant hypertension12,13,17,22,26,28,39–46 that consistently show blood- function worsened in some patients, possibly because of pressure-lowering eff ects after this procedure in patients with resistant hypertension, but progression before hypertensive kidney disease, con- mostly up to 12 months. Longer-term follow-up has been limited to the previous tinued renal injury due to hypertension, or changes in Symplicity HTN-1 report in 18 patients at 24 months.12 No notable safety concerns were diuretic use.25 An analysis of 88 patients showed that raised in these trials, and in a trial done to assess renal haemodynamics and renal function percutaneous RDN did not negatively aff ect renal after catheter-based RDN showed no adverse eff ects on glomerular fi ltration rate or renal 26 function, as measured by cystatin C eGFR. RDN has artery structure in 88 patients after 6 months.26 been suggested as a useful approach to manage high-risk patients with hypertension and kidney disease.27 The Interpretation feasibility of using RDN in patients with chronic kidney After follow-up of 36 months in patients with resistant hypertension who had previously disease and treatment-resistant hypertension was tested not responded to multiple (mean fi ve) antihypertensive medications, RDN was associated in 15 patients, in whom blood pressure was signifi cantly with a signifi cant and sustained reduction in blood pressure without serious adverse reduced with no decline in eGFR over 12 months.28 This eff ects. RDN seems to off er a novel and complementary approach for patients in whom is an active area of research and clearly requires longer pharmacological options seem exhausted. Adequately powered clinical trials with follow-up of larger numbers of patients. long-term follow-up are warranted to confi rm these results in larger numbers of patients. Our study has several limitations. The fi rst is that only a subset of the original 153 patients consented to 36 months of follow-up and of these not all had full patients were receiving an aldosterone-receptor antagon- data available at the fi nal follow-up visit. Nevertheless, ist at baseline. These drugs are thought to be useful as although small, our analysis is among the largest and the add-on therapy in patients with treatment-resistant only one so far to report follow-up to 36 months. Selection hypertension and their use should be considered before bias cannot be excluded, but long-term safety and RDN.17,30,31 Caution should, however, be applied. In durability of blood-pressure reduction at 36 months was position papers from the European Society of Cardiology seen in most patients. and the European Society of Hypertension, concerns We did not include a control group in this proof-of- have been raised about the long-term safety of concept trial and, therefore, issues of regression to the aldosterone-receptor antagonists, especially in patients mean, placebo, and Hawthorne eff ects might aff ect the with reduced renal function and existing blockade of the fi ndings even late after the procedure. For example, renin-angiotensin system.20,32 blood pressure might drop simply because patients have Another limitation is the lack of data to assess the improved adherence to lifestyle changes or medication eff ects of medication dose or drug changes within a regimens while being observed in a study. The long-term medication class. These data were not collected beyond blood-pressure eff ects, however, diff er from the expected 12 months and, therefore, are not reported. Thus, we disease course in patients with treatment-resistant draw no conclusions about the potential eff ects of hypertension, which is progressive worsening of blood- antihypertensive medication changes on outcomes or pressure control.6,29 whether the procedure alters the ongoing need for All patients might not have been receiving optimum certain antihypertensive medication. The role of anti- medical treatment before RDN. Specifi cally, only 28% of hypertensive medications or other interventions in the www.thelancet.com Vol 383 February 15, 2014 627 Downloaded for Anonymous User (n/a) at West Virginia School of Osteopathic Medicine from ClinicalKey.com by Elsevier on August 17, 2021. For personal use only. No other uses without permission. Copyright ©2021. Elsevier Inc. All rights reserved. Articles

lowering of blood pressure cannot be ruled out in these 4 Dzau VJ, Antman EM, Black HR, et al. The cardiovascular disease patients. Ambulatory monitoring of blood pressure was continuum validated: clinical evidence of improved patient outcomes: part I: pathophysiology and clinical trial evidence (risk not included in the follow-up period. factors through stable coronary artery disease). Circulation 2006; Finally, these results were obtained in a select group of 114: 2850–70. patients who met the study inclusion and exclusion 5 Sarafi dis PA, Bakris GL. Resistant hypertension: an overview of evaluation and treatment. J Am Coll Cardiol 2008; 52: 1749–57. criteria and were willing to participate in a research study 6 Calhoun DA, Jones D, Textor S, et al, for the American Heart and thus may not be broadly applicable to real-world Association Professional Education Committee. Resistant hypertensive patients. The Global SYMPLICITY registry hypertension: diagnosis, evaluation, and treatment: A scientifi c statement from the American Heart Association Professional aims to accrue up to 5000 hypertensive patients from Education Committee of the Council for High Blood Pressure practice to better assess this issue.33 Research. Circulation 2008; 117: e510–26. Overactivity of the sympathetic nervous system plays 7 Waeber B, Feihl F. Assessment of drug compliance in patients with high blood pressure resistant to antihypertensive therapy. an important role in the development of hypertension. EuroIntervention 2013; 9 (suppl R): R29–34. RDN is a straightforward, minimally invasive per- 8 Sobotka PA, Mahfoud F, Schlaich MP, Hoppe UC, Böhm M, cutaneous method of denervation (panel). We found that Krum H. Sympatho-renal axis in chronic disease. Clin Res Cardiol this procedure was associated with substantial lowering 2011; 100: 1049–57. 9 Esler M, Jennings G, Korner P, et al. Assessment of human of blood pressure in patients with treatment-resistant sympathetic nervous system activity from measurements of hypertension, could be performed without any major norepinephrine turnover. Hypertension 1988; 11: 3–20. safety issues, and that the eff ects persisted up to 10 Schlaich MP, Lambert E, Kaye DM, et al. Sympathetic augmentation in hypertension: role of nerve fi ring, norepinephrine reuptake, and 36 months. The blood-pressure-lowering eff ects were not angiotensin neuromodulation. Hypertension 2004; 43: 169–75. altered by age, baseline renal function, or diabetes status. 11 Esler MD, Krum H, Sobotka PA, Schlaich MP, Schmieder RE, We suggest that durable responses after RDN will Böhm M, for the Symplicity HTN-2 Investigators. Renal sympathetic denervation in patients with treatment-resistant eventually be possible in most if not all patients with hypertension (The Symplicity HTN-2 Trial): a randomised treatment-resistant hypertension. controlled trial. Lancet 2010; 376: 1903–09. Contributors 12 Symplicity HTN-1 Investigators. Catheter-based renal sympathetic denervation for resistant hypertension: durability of blood pressure HK was the principal investigator and wrote the paper. All authors reduction out to 24 months. Hypertension 2011; 57: 911–17. participated in data collection and interpretation. MPS, MB, FM, KR-S, 13 Esler MD, Krum H, Schlaich M, Schmieder RE, Böhm M, RK, and MDE provided critical revisions of the drafts. PAS contributed Sobotka PA, for the Symplicity HTN-2 Investigators. Renal to the original hypothesis, trial execution, data analysis and sympathetic denervation for treatment of drug-resistant interpretation, and manuscript writing and review. hypertension: one-year results from the Symplicity HTN-2 Confl icts of interest randomized, controlled trial. Circulation 2012; 126: 2976–82. All authors received a research grant from Medtronic to undertake the 14 Schlaich MP, Sobotka PA, Krum H, Lambert E, Esler MD. Renal Symplicity HTN-1 clinical trial. HK, MPS, and MDE have received sympathetic-nerve ablation for uncontrolled hypertension. N Engl J Med 2009; 361: 932–34. honoraria and travel support from Medtronic. MPS and MDE are supported by an NHMRC Senior Research Fellowship. PAS receives 15 Hering D, Lambert EA, Marusic P, et al. Substantial reduction in single sympathetic nerve fi ring after renal denervation in patients royalties related to sales of the Medtronic renal denervation catheter, is a with resistant hypertension. Hypertension 2013; 61: 457–64. consultant for Medtronic, Abbott Ventures, and Rox Medical, and is an 16 Ewen S, Ukena C, Böhm M, Mahfoud F. Percutaneous renal employee of Cibiem. MB and FM are supported by the Deutsche denervation: new treatment option for resistant hypertension and Forschungsgemeinschaft (KFO 196) and received speaker honoraria more? Heart 2013; 99: 1129–34. from Medtronic, St Jude Medical, and Cordis, and scientifi c support 17 Krum H, Schlaich M, Whitbourn R, et al. Catheter-based renal from Medtronic, St Jude Medical, Vessix, and Recor. FM is supported by sympathetic denervation for resistant hypertension: a multicentre Deutsche Hochdruckliga and Deutsche Gesellschaft für Kardiologie. safety and proof-of-principle cohort study. Lancet 2009; 373: 1275–81. KR-S is a consultant for Medtronic, BSC, CardioSonic, and CiBiem. 18 Schlaich MP, Hering D, Sobotka P, et al. Eff ects of renal Acknowledgments denervation on sympathetic activation, blood pressure, and glucose Ardian LLC was the sponsor of this study until January, 2011, when it metabolism in patients with resistant hypertension. Front Physiol was acquired by Medtronic, who continued sponsorship. We wish to 2012; 3: 10. thank Colleen Gilbert for liaising with Medtronic and providing editorial 19 Schmieder RE, Redon J, Grassi G, et al. ESH position paper: renal support, Martin Fahy and Minglei Liu for statistical analysis, denervation—an interventional therapy of resistant hypertension. J Hypertens 2012; 30: 837–41. Christina Hawley for data management support, and Sidney A Cohen 20 Mahfoud F, Lüscher TF, Andersson B, et al. Expert consensus for review of and suggestions to improve the text. document from the European Society of Cardiology on References catheter-based renal denervation. Eur Heart J 2013; 34: 2149–57. 1 Ezzati M, Lopez AD, Rodgers A, Vander Hoorn S, Murray CJ, and 21 Lewington S, Clarke R, Qizilbash N, Peto R, Collins R, Prospective the Comparative Risk Assessment Collaborating Group. Selected Studies C, for the Prospective Studies Collaboration. Age-specifi c major risk factors and global and regional burden of disease. Lancet relevance of usual blood pressure to vascular mortality: 2002; 360: 1347–60. a meta-analysis of individual data for one million adults in 2 Chobanian AV, Bakris GL, Black HR, et al, for the Joint National 61 prospective studies. Lancet 2002; 360: 1903–13. Committee on Prevention, Detection, Evaluation, and Treatment of 22 Mahfoud F, Ukena C, Schmieder RE, et al. Ambulatory blood High Blood Pressure; National Heart, Lung, and Blood Institute, pressure changes after renal sympathetic denervation in patients and the National High Blood Pressure Education Program with resistant hypertension. Circulation 2013; 128: 132–40. Coordinating Committee. Seventh report of the joint national 23 Kaltenbach B, Id D, Franke JC, et al. Renal artery stenosis after committee on prevention, detection, evaluation, and treatment of renal sympathetic denervation. J Am Coll Cardiol 2012; 60: 2694–95. high blood pressure. Hypertension 2003; 42: 1206–52. 24 Rippy MK, Zarins D, Barman NC, Wu A, Duncan KL, Zarins CK. 3 Kearney PM, Whelton M, Reynolds K, Muntner P, Whelton PK, Catheter-based renal sympathetic denervation: chronic preclinical He J. Global burden of hypertension: analysis of worldwide data. evidence for renal artery safety. Clin Res Cardiol 2011; 100: 1095–101. Lancet 2005; 365: 217–23.

628 www.thelancet.com Vol 383 February 15, 2014 Downloaded for Anonymous User (n/a) at West Virginia School of Osteopathic Medicine from ClinicalKey.com by Elsevier on August 17, 2021. For personal use only. No other uses without permission. Copyright ©2021. Elsevier Inc. All rights reserved. Articles

25 Bakris GL. Treatment of stage I hypertension and development of 37 Curtis JJ, Lucas BA, Kotchen TA, Luke RG. Surgical therapy for renal dysfunction. J Hum Hypertens 2001; 15: 81–84. persistent hypertension after renal transplantation. Transplantation 26 Mahfoud F, Cremers B, Janker J, et al. Renal hemodynamics and 1981; 31: 125–28. renal function after catheter-based renal sympathetic denervation 38 Onesti G, Kim KE, Greco JA, del Guercio ET, Fernandes M, in patients with resistant hypertension. Hypertension 2012; Swartz C. Blood pressure regulation in end-stage renal disease and 60: 419–24. anephric man. Circ Res 1975; 36 (suppl 1): 145–52. 27 Hering D, Esler MD, Schlaich MP. Chronic kidney disease: role of 39 Mahfoud F, Schlaich M, Kindermann I, et al. Eff ect of renal sympathetic nervous system activation and potential benefi ts of sympathetic denervation on glucose metabolism in patients with renal denervation. EuroIntervention 2013; 9 (suppl R): R127–35. resistant hypertension: a pilot study. Circulation 2011; 123: 1940–46. 28 Hering D, Mahfoud F, Walton AS, et al. Renal denervation in 40 Witkowski A, Prejbisz A, Florczak E, et al. Eff ects of renal moderate to severe CKD. J Am Soc Nephrol 2012; 23: 1250–57. sympathetic denervation on blood pressure, sleep apnea course, 29 Segura J, de la Sierra A, Ruilope LM. Detection and treatment of and glycemic control in patients with resistant hypertension and resistant hypertension. Curr Hypertens Rep 2010; 12: 325–30. sleep apnea. Hypertension 2011; 58: 559–65. 30 Chapman N, Dobson J, Wilson S, et al, for the Anglo-Scandinavian 41 Ukena C, Mahfoud F, Kindermann I, et al. Cardiorespiratory Cardiac Outcomes Trial Investigators. Eff ect of spironolactone on response to exercise after renal sympathetic denervation in patients blood pressure in subjects with resistant hypertension. Hypertension with resistant hypertension. J Am Coll Cardiol 2011; 58: 1176–82. 2007; 49: 839–45. 42 Ott C, Janka R, Schmid A, et al. Vascular and renal hemodynamic 31 Václavík J, Sedlák R, Plachy M, et al. Addition of spironolactone changes after renal denervation. Clin J Am Soc Nephrol 2013; in patients with resistant arterial hypertension (ASPIRANT): 8: 1195–201. a randomized, double-blind, placebo-controlled trial. Hypertension 43 Ormiston JA, Watson T, van Pelt N, et al. Renal denervation for 2011; 57: 1069–75. resistant hypertension using an irrigated radiofrequency balloon: 32 Schmieder RE, Redon J, Grassi G, et al. Updated ESH position 12-month results from the Renal Hypertension Ablation System paper on interventional therapy of resistant hypertension. (RHAS) trial. EuroIntervention 2013; 9: 70–74. EuroIntervention 2013; 9 (suppl R): R58–66. 44 Worthley SG, Tsioufi s CP, Worthley MI, et al. Safety and effi cacy of 33 Böhm M, Mahfoud F, Ukena C, et al. Rationale and design of a a multi-electrode renal sympathetic denervation system in resistant large registry on renal denervation: the Global SYMPLICITY hypertension: the EnligHTN I trial. Eur Heart J 2013; 34: 2132–40. registry. EuroIntervention 2013; 9: 484–92. 45 Schmid A, Ditting T, Sobotka PA, et al. Does renal artery supply 34 Esler M, Ferrier C, Lambert G, Eisenhofer G, Cox H, Jennings G. indicate treatment success of renal denervation? Biochemical evidence of sympathetic hyperactivity in human Cardiovasc Intervent Radiol 2013; 36: 987–91. hypertension. Hypertension 1991; 17 (suppl): III29–35. 46 Hering D, Lambert EA, Marusic P, et al. Renal nerve ablation 35 Smithwick RH, Thompson JE. Splanchnicectomy for essential reduces augmentation index in patients with resistant hypertension; results in 1,266 cases. J Am Med Assoc 1953; hypertension. J Hypertens 2013; published online May 21. 152: 1501–04. DOI:10.1097/HJH.0b013e3283622e58. 36 Cannon B. The eff ects of progressive sympathectomy on blood pressure. Am J Physiol 1931; 97: 592–96.

www.thelancet.com Vol 383 February 15, 2014 629 Downloaded for Anonymous User (n/a) at West Virginia School of Osteopathic Medicine from ClinicalKey.com by Elsevier on August 17, 2021. For personal use only. No other uses without permission. Copyright ©2021. Elsevier Inc. All rights reserved.