BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
Therapy discontinuation or substitution in patients with cardiovascular disease, switching among different products of the same off-patent active substance: a “real-world” retrospective cohort study
For peer review only Journal: BMJ Open
Manuscript ID bmjopen-2016-012003
Article Type: Research
Date Submitted by the Author: 21-Mar-2016
Complete List of Authors: Degli Esposti, Luca; CliCon S.r.l, Health, Economics and Outcomes Research, Sangiorgi, Diego; CliCon S.r.l. Health, Economics and Outcomes Research Buda, Stefano; CliCon S.r.l. Health, Economics and Outcomes Research Degli Esposti, Ezio; CliCon S.r.l. Health, Economics and Outcomes Research Scaglione, Francesco ; University of Milan, Medical Biotechnology and Translational Medicine
Primary Subject Cardiovascular medicine Heading:
Secondary Subject Heading: Public health http://bmjopen.bmj.com/ Keywords: Simvastatin, Amlodipine, Ramipril, Switching, Adherence to treatment
on September 24, 2021 by guest. Protected copyright.
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 1 of 23 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 5 Therapy discontinuation or substitution in patients with cardiovascular disease, 6 7 switching among different products of the same off-patent active substance: a “real- 8 9 world” retrospective cohort study 10 11 12 13 14 Luca Degli Esposti,1 Diego Sangiorgi,1 Stefano Buda,1 Ezio Degli Esposti,1 Francesco 15 For peer review only 16 2 17 Scaglione 18 19 20 21 22 1CliCon S.r.l. Health, Economics and Outcomes Research, Ravenna, Italy 23 24 2 25 Department of Medical Biotechnology and Translational Medicine, University of Milan, 26 27 Milan, Italy 28 29 30 31 32 Correspondence to 33
34 http://bmjopen.bmj.com/ 35 Luca Degli Esposti, EconD., CliCon S.r.l, Health, Economics and Outcomes Research, Via 36 37 Salara, 36 48100 Ravenna, Italy 38 39 40 Tel +39 544 38393 Fax +39 544 212699. [email protected] 41
42 on September 24, 2021 by guest. Protected copyright. 43 44 45 Keywords 46 47 48 Simvastatin, Amlodipine, Ramipril, Switching, Adherence to treatment 49 50 51 52 53 Word count: 2,452 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 2 of 23 BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 ABSTRACT 5 6 Objective: The present study investigated the effects of switching to different products of the 7 8 same off patent active substance (brand name or generic) on therapy discontinuation or 9 10 11 substitution with another molecule of the same class, in patients with cardiovascular disease 12 13 treated with statins and antihypertensives in clinical practice. 14 15 For peer review only 16 Design: A retrospective cohort study in a “real world” setting. 17 18 Setting: Analysis of data performed by integrating administrative databases from three 19 20 21 Italian Local Health Units located in three different regions with a total population of about 22 23 21 million. 24 25 Participants: All patients aged ≥18 years with at least one prescription of simvastatin, 26 27 st st 28 ramipril or amlodipine in the period January 1 to December 31 2010 were included and 29 30 followed up for two years. 31 32 33 Main outcome measures: Prescription refills occurring during follow up were evaluated.
34 http://bmjopen.bmj.com/ 35 Frequency of discontinuation of therapy or substitution with another molecule of the same 36 37 class (for example, from simvastatin to a different statin) during follow up was identified. 38 39 40 Results: During follow up, therapy discontinuation or substitution was found to be more 41
42 frequent in patients switching to a different product of the same active substance compared on September 24, 2021 by guest. Protected copyright. 43 44 with non switching patients (11.5% vs 10.8% and 22.2% vs 20.8% [p=0.002], respectively, 45 46 in the simvastatin group; 7.6% vs 7.8% and 19.2% vs 17.0% [p<0.001], respectively, in the 47 48 49 ramipril group; 4.0% vs 3.5% and 24.6% vs 22.7% [p<0.001], respectively, in the amlodipine 50 51 group). These findings were confirmed by multivariate analysis. 52 53 54 Conclusions: Switches among products of the same active substance are quite common in 55 56 patients with cardiovascular disease. Our study suggests that switching may expose patients 57 58 to a higher risk of therapy discontinuation or substitution. 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 3 of 23 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 5 6 7 STRENGTHS AND LIMITATIONS OF THIS STUDY 8 9 • This study, in a “real world” setting, is one of only a few studies to investigate 10 11 12 clinical differences related to switching among different products of the same active 13 14 substance in the cardiovascular setting. Until now, most research has focused only on 15 For peer review only 16 comparing brand name and generic drugs. 17 18 19 • Sample size was relatively limited, and although we used three healthcare databases 20 21 from regions with a total population of about 21 million, large studies are needed to 22 23 confirm and to enhance the generalizability of the findings, and in different 24 25 populations. 26 27 28 • In common with other retrospective, observational studies, reasons for switch, non 29 30 adherence or discontinuation of treatment were not retrievable from the data set. 31 32 33
34 http://bmjopen.bmj.com/ 35 36 37 38 39 40 41
42 on September 24, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 4 of 23 BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 INTRODUCTION 5 6 7 Cardiovascular diseases (CVD) are the leading cause of death worldwide, accounting for 8 9 approximately one third of all deaths.[1] Combination therapy with antihypertensive drugs 10 11 12 and serum cholesterol lowering drugs is effective in prevention, and it is estimated that a high 13 14 level of adherence to treatment will reduce the risk of CVD by approximately 80%.[2] A 15 For peer review only 16 number of studies have demonstrated that patients often discontinue long term treatment or 17 18 take less than prescribed, and that such non adherence reduces the potential preventive 19 20 benefits.[3] 21 22 23 Many reasons contribute to patient non adherence to medical therapy, such as ageing, 24 25 comorbidities, polypharmacy, poor relationship between patient and physician, poor memory, 26 27 28 and patients’ low perception of disease severity.[4] On the other hand, it is unlikely that side 29 30 effects are the main cause of poor adherence to preventive treatment, as there seems to be 31 32 little direct relationship between adherence and drug class.[3] Furthermore, some studies 33 34 have demonstrated that switching between different products of the same active substance http://bmjopen.bmj.com/ 35 36 can have an impact on adherence to medication, because variation in packaging and pill 37 38 39 appearance may reduce adherence, especially for chronic diseases.[5, 6] 40 41 There is a perception among patients and physicians alike that frequent changes between
42 on September 24, 2021 by guest. Protected copyright. 43 44 branded and unbranded products (as well as between generics), all containing the same active 45 46 substance, and especially if patients are older and on multi drug regimens, may cause patients 47 48 to become anxious when the appearance of their drugs changes.[7 9] This can lead to an 49 50 increased risk of patients making mistakes or double medicating, which flows on to increased 51 52 53 drug non adherence.[10, 11] 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 5 of 23 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 Few studies have investigated clinical differences related to switching among different 5 6 products of the same active substance in the cardiovascular setting. Until now, most research 7 8 has focused only on comparing brand name and generic drugs.[12 16] 9 10 11 The aim of the present study was to investigate the effects of switching to different 12 13 products of the same off patent active substance (brand name or generic) on therapy 14 15 For peer review only 16 discontinuation or substitution with another molecule of the same class, in patients with CVD 17 18 treated with statins and antihypertensives in clinical practice. 19 20 21 A version of this article has been previously published as a journal supplement in the 22 23 Italian language.[17] 24 25 26 METHODS 27 28 29 Data collection 30 31 The data used for the analysis were obtained from the administrative databases of three Local 32 33 Health Units (LHUs), including a total population of about two million health assisted 34 http://bmjopen.bmj.com/ 35 36 individuals, in the Italian regions of Lombardy, Lazio and Campania. We analyzed the 37 38 following archives: Assisted Subjects’ Database, containing the personal data of patients; 39 40 Medication Prescription database, containing all the information relating to individual 41
42 prescriptions, such as the International Nonproprietary Names (INN) for pharmaceutical on September 24, 2021 by guest. Protected copyright. 43 44 45 substances, the Anatomical Therapeutic Chemical (ATC) code of the prescribed drug, the 46 47 number of packages, the number of units per package, the dose, the cost per unit and the date 48 49 of the prescription; Hospital Discharge Database (SDO), containing information on each 50 51 hospital discharge, in particular the date of admission and discharge, primary and secondary 52 53 diagnoses coded according to the International Classification of Diseases, Ninth Revision, 54 55 56 Clinical Modification (ICD-9-CM). The patient code in each database allowed electronic 57 58 linking among all databases. To guarantee patient privacy, this patient code was transcoded 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 6 of 23 BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 into an anonymous univocal numeric code. No identifiers related to patients were provided to 5 6 the researchers. According to the Italian law for confidentiality of data[18] the study was 7 8 notified to the Ethic Committees of each LHU. 9 10 11 Cohort definition 12 13 14 The study was a retrospective cohort study including all patients aged ≥18 years that, 15 For peer review only 16 between January 1st 2010 and December 31st 2010 (enrollment period), had at least one 17 18 prescription of simvastatin [ATC code: C10AA01], ramipril [ATC code: C09AA05] or 19 20 21 amlodipine [ATC code: C08CA01]. The date of enrollment was defined as the earliest date 22 23 within the enrollment period in which the patient had the last switch of medication 24 25 (switchers) or the last prescription in the case of a patient continuing with the same 26 27 medication (non switchers). Starting from this date, the individual patient was followed for 28 29 two years (follow up period). Switchers were defined as those patients who switched among 30 31 32 different products of the same off patent active substance (i.e., from brand name to generic, 33
34 from generic to brand name, or from generic to another generic). The changes in dose and http://bmjopen.bmj.com/ 35 36 dosage form were not accounted for during switching. The definitions of the patient cohorts 37 38 are shown in Figure 1. Data on baseline characteristics, including demographics, 39 40 41 cardiovascular risk factors and polytherapies, were collected; specifically, polytherapies were
42 on September 24, 2021 by guest. Protected copyright. 43 identified as more than one cardiovascular medication at the date of enrollment. The data on 44 45 drug prescriptions and hospitalizations that occurred during the 12 months preceding the date 46 47 of enrollment were analyzed (characterization period). 48 49 50 Only patients with at least one prescription of the index drug in the previous 12 months 51 52 were included (to capture patients who could have made a change in therapy) and with at 53 54 least two prescriptions at follow up (to include patients with continuity of treatment). Patients 55 56 57 treated with fixed combinations of the molecules under consideration (ramipril and diuretics 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 7 of 23 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 [ATC code: C09BA05], perindopril and amlodipine [ATC code: C09BB04], ramipril and 5 6 felodipine [ATC code: C09BB05], olmesartan and amlodipine [ATC code: C09DB02], 7 8 simvastatin and ezetimibe [ATC code: C10BA02]) were excluded. 9 10 11 Patients transferred to another LHU during the follow up period were excluded from the 12 13 analysis. 14 15 For peer review only 16 Study population 17 18 19 Cardiovascular risk 20 21 22 Patients were classified as high cardiovascular risk if they had cardiovascular treatment or 23 24 hospitalization for diabetes. Hospitalizations related to diabetes were identified by ICD 9 CM 25 26 code: 250 (primary discharge reasons); and/or cardiovascular risk factors (previous 27 28 hospitalization for Acute Cardiac Ischemia [ICD 9 CM: 411], Angina Pectoris [ICD 9 CM: 29 30 413] Chronic Cardiac Ischemia [ICD 9 CM: 414]; Cerebral Hemorrhage [ICD 9 CM: 431]; 31 32 33 Cerebral Artery Occlusion [ICD 9 CM: 434]; Transient Cerebral Ischemia [ICD 9 CM: 435];
34 http://bmjopen.bmj.com/ 35 Cerebral Circulatory Disorders [ICD 9 CM: 436]; Atherosclerosis [ICD 9 CM: 440]; Other 36 37 Peripheral Vascular Disease [ICD 9 CM: 443]; Chronic Renal Failure [ICD 9 CM: 585]; 38 39 Coronary Angioplasty [ICD 9 CM procedure: 0066, 360]); and/or at least two prescriptions of 40 41 antidiabetic drugs [ATC code: A10]. All other patients were classified as at moderate
42 on September 24, 2021 by guest. Protected copyright. 43 44 cardiovascular risk. 45 46 Drug treatments 47 48 49 Patients were also characterized by strategy of treatment at baseline with lipid lowering drugs 50 51 (ATC code: C10) and antihypertensive drugs (ATC codes: C02, C03, C07, C08, C09). 52 53 54 Data analysis at follow up 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 8 of 23 BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 During the follow up period, the discontinuation or the first substitution of therapy was 5 6 identified. Discontinuation of therapy was defined as the absence of prescriptions of the same 7 8 therapeutic class (ATC group) as the index molecule in the last quarter of observation. A 9 10 11 substitution of therapy was defined as a change to a different active substance of the same 12 13 therapeutic class (ATC group) (i.e., switching from simvastatin to a different statin). A switch 14 15 among differentFor products peer of the same active review substance was identified only by INN. 16 17 18 Statistical analysis 19 20 21 Continuous variables were reported as mean ± standard deviation (SD) and compared using 22 23 the Student's t test; categorical variables were reported as absolute numbers and percentages 24 25 and compared using the chi square test. 26 27 28 Discontinuations of therapy and substitution with another molecule of the same class 29 30 were analyzed by multivariate analysis using Cox proportional hazards models; covariates 31 32 considered in the models were: age, male sex, high cardiovascular risk, cardiovascular 33 34 treatments, change of formulation in the period of characterization. http://bmjopen.bmj.com/ 35 36 37 The analysis of Schoenfeld residuals (scaled and unscaled) was conducted to assess the 38 39 proportionality of risk. 40 41
42 p Values <0.05 were considered statistically significant. All analyses were performed on September 24, 2021 by guest. Protected copyright. 43 44 using STATA 12.0 SE. 45 46 47 RESULTS 48 49 50 Simvastatin 51 52 53 A total of 38,183 patients treated with simvastatin, 17,642 male (46%), mean age 68.3±10.7 54 55 years, were included in the analysis. A total of 9,392 (25%) patients were classified as at high 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 9 of 23 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 cardiovascular risk, while 30,467 (80%) received concomitant cardiovascular treatments 5 6 (Table 1). 7 8 Table 1 Characteristics of the enrolled patients 9 10 11 Simvastatin Ramipril Amlodipine 12 13 Non- p Non- p Non- p 14 15 switchersFor Switchers peer Value reviewswitchers Switchers onlyValue switchers Switchers Value 16 17 N (%) 23,180 (61) 15,003 (39) 22,799 (71) 9,312 (29) 26,823 (72) 10,644 (28) 18 19 20 Age, y ±SD 68.5 ± 10.7 68.0 ± 10.6 <0.001 67.0 ± 12.7 66.8 ± 13.0 0.203 68.3 ± 11.7 68.1 ± 11.8 0.137 21 22 Male sex, N (%) 10,469 (45.2) 7,173 (47.8) <0.001 12,876 5,617 (60.3) <0.001 14,126 6,213 (58.4) <0.001 23 (56.5) (52.7) 24 25 26 High CVR, N (%) 5,494 (23.7) 3,898 (26.0) <0.001 4,873 (21.4) 2,025 (21.7) 0.470 4,983 (18.6) 2,143 (20.1) <0.001 27 28 Additional 18,539 (80.0) 11,928 0.265 17,868 7,393 (79.4) 0.044 23,793 9,588 (90.1) <0.001 29 treatments, N (%) (79.5) (78.4) (88.7) 30 31 32 CVR, cardiovascular risk; SD, standard deviation. 33
34 http://bmjopen.bmj.com/ 35 36 37 Switches among different products occurred in 39% of patients treated with simvastatin. With 38 39 regard to switchers, a little over half carried out one switch only during the characterization 40 41 period, with 8% having four switches or more (Table 2).
42 on September 24, 2021 by guest. Protected copyright. 43 44 45 46 47 Table 2 Annual frequency of switches 48 49 50 Simvastatin Ramipril Amlodipine 51 52 53 Switches (N) Patients (%) Switches (N) Patients (%) Switches (N) Patients (%) 54 55 1 7,842 (52) 1 5,112 (55) 1 5,710 (54) 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 10 of 23 BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 2 4,062 (27) 2 2,501 (27) 2 2,871 (27) 5 6 7 3 1,917 (13) 3 1,116 (12) 3 1,300 (12) 8 9 4 805 (5) 4 413 (4) 4 563 (5) 10 11 ≥5 377 (3) ≥5 170 (2) ≥5 200 (2) 12 13 14 15 For peer review only 16 17 In the follow up period, 4,232 (11%) patients undertook a therapy substitution with another 18 19 molecule; a significantly lower percentage of substitution was observed in the group that did 20 21 not switch to a different product of the same active substance (Table 3). 22 23 24 25 26 Table 3 Therapy discontinuation and substitution in switcher and non switcher patients 27 28 29 Simvastatin Ramipril Amlodipine 30 31 Switch Switch Switch 32 33 No Yes p No Yes p No Yes p 34 http://bmjopen.bmj.com/ 35 Value Value Value 36 37 Patients, N (%) 23,180 15,003 22,799 (71) 9,312 (29) 26,823 (72) 10,644 (28) 38 39 (61) (39) 40 41 Substitution, N (%) 2,506 1,726 0.037 1,785 (7.8) 711 (7.6) 0.571 943 (3.5) 426 (4.0) 0.026
42 (10.8) (11.5) on September 24, 2021 by guest. Protected copyright. 43 44 45 Discontinuation, N (%) 4,826 3,327 0.002 3,885 (17.0) 1,792 (19.2) <0.001 6,089 (22.7) 2,618 (24.6) <0.001 46 (20.8) (22.2) 47 48 49 50 51 In the same period 8,153 (21%) patients discontinued treatment; a significantly lower 52 53 54 percentage of discontinuation was observed for non switching patients (Table 3). These 55 56 findings were partially confirmed by multivariate analysis (Table 4): the group that switched 57 58 to a different product of the same active substance showed a higher probability of 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 11 of 23 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 discontinuation (hazard ratio [HR]=1.087, p<0.001) and a higher, but not significant, 5 6 probability of substitution of therapy (HR=1.059, p=0.068). 7 8 9 Table 4 Multivariate analysis of predictors of risk of therapy discontinuation and substitution 10 11 12 Variable Simvastatin 13 14 Substitution Discontinuation 15 For peer review only 16 17 HR 95% CI p Value HR 95% CI p Value 18 19 Age 0.993 0.990 0.995 <0.001 1.001 0.999 1.004 0.194 20 21 22 Male sex 1.116 1.050 1.186 <0.001 0.921 0.882 0.963 <0.001 23 24 High CVR 1.107 1.034 1.185 0.004 0.962 0.914 1.013 0.140 25 26 Additional CV treatments 1.064 0.984 1.152 0.121 0.821 0.778 0.867 <0.001 27 28 29 Switch to different product of 30 the same substance 1.059 0.996 1.126 0.068 1.087 1.040 1.136 <0.001 31 32 Ramipril 33
34 http://bmjopen.bmj.com/ 35 Substitution Discontinuation 36 37 HR 95% CI p Value HR 95% CI p Value 38 39 40 Age 1.003 1.000 1.006 0.072 1.005 1.002 1.007 <0.001 41
42 Male sex 0.938 0.865 1.017 0.119 0.950 0.900 1.002 0.058 on September 24, 2021 by guest. Protected copyright. 43 44 High CVR 1.158 1.057 1.269 0.002 0.866 0.811 0.926 <0.001 45 46 47 Additional CV treatments 1.521 1.360 1.700 <0.001 0.806 0.757 0.857 <0.001 48 49 Switch to different product of 50 the same substance 0.973 0.892 1.062 0.540 1.163 1.100 1.230 <0.001 51 52 53 Amlodipine 54 55 Substitution Discontinuation 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 12 of 23 BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 HR 95% CI p Value HR 95% CI p Value 5 6 7 Age 1.019 1.013 1.024 <0.001 1.006 1.004 1.008 <0.001 8 9 Male sex 0.807 0.719 0.907 <0.001 0.870 0.834 0.909 <0.001 10 11 High CVR 1.247 1.091 1.424 0.001 0.879 0.831 0.929 <0.001 12 13 14 Additional CV treatments 1.935 1.514 2.474 <0.001 0.959 0.897 1.025 0.215 15 For peer review only 16 Switch to different product of 17 18 the same substance 1.179 1.043 1.333 0.008 1.124 1.074 1.177 <0.001 19 20 CI, confidence interval; CV, cardiovascular; CVR, cardiovascular risk; HR, hazard ratio. 21 22 23 24 25 Ramipril 26 27 A total of 32,111 patients treated with ramipril, 18,493 male (58%), mean age 66.9±12.8 28 29 30 years, were included in the analysis. Of these, 6,898 (21%) patients were classified as at high 31 32 cardiovascular risk, while 25,261 (79%) were receiving additional cardiovascular treatments 33
34 (Table 1). Switches among different products occurred in 29% of patients treated with http://bmjopen.bmj.com/ 35 36 ramipril. With regard to switchers, again, a little over half (55%) carried out one switch only, 37 38 during the characterization period, and few (6%) had four switches or more (Table 2). In the 39 40 41 follow up period, 2,496 (8%) patients undertook a therapy substitution to another molecule;
42 on September 24, 2021 by guest. Protected copyright. 43 no trend towards a lower percentage of substitution was observed in the group that did not 44 45 switch to a different product of the same active substance (Table 3). In the same period 5,677 46 47 (18%) patients discontinued treatment, with a significant difference in favor of patients not 48 49 50 switching to a different product of the same active substance (Table 3). These findings were 51 52 confirmed by multivariate analysis (Table 4): there was essentially no difference between 53 54 groups in terms of probability of substitution (HR=0.973, p=0.540), while the non switching 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 13 of 23 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 group showed a significantly lower probability of discontinuation of therapy (HR=1.163, 5 6 p<0.001). 7 8 9 Amlodipine 10 11 12 A total of 37,467 patients treated with amlodipine, 20,339 male (54%), mean age 68.2±11.7 13 14 years, were included in the analysis. Of these, 7,126 (19%) patients were classified as at high 15 For peer review only 16 cardiovascular risk, while 33,381 (89%) were receiving additional cardiovascular treatments 17 18 (Table 1). Switches among different products occurred in 28% of patients treated with 19 20 21 amlodipine. With regard to switchers, just over half (54%) carried out one switch only during 22 23 the characterization period, and 7% had four switches or more (Table 2). In the follow up 24 25 period, 1,369 (4%) patients undertook a therapy substitution to another molecule; a 26 27 significantly lower percentage of substitution was observed in the group that did not switch to 28 29 a different product of the same active substance (Table 3). In the same period 8,707 (23%) 30 31 32 patients discontinued treatment; a significantly lower probability of discontinuation was 33
34 observed for patients not switching to another product of the same active substance (Table 3). http://bmjopen.bmj.com/ 35 36 These findings were confirmed by multivariate analysis (Table 4): the switcher group showed 37 38 a higher probability of discontinuation (HR=1.124, p<0.001) and substitution of therapy 39 40 41 (HR=1.179, p=0.008).
42 on September 24, 2021 by guest. Protected copyright. 43 DISCUSSION 44 45 46 In accordance with previous studies,[19 22] this retrospective analysis in a “real world” 47 48 49 setting shows that age, gender, cardiovascular risk, and polypharmacy affect discontinuation 50 51 of therapy. A number of factors may interact to affect adherence to therapies for chronic 52 53 conditions. These have been categorized by the World Health Organization as social and 54 55 economic related factors, health system/health care team related factors, condition related 56 57 factors, and patient related factors.[11] As poor adherence has a significant negative impact 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 14 of 23 BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 on health outcomes and healthcare costs, and imposes a substantial burden on patients and 5 6 health systems, “increasing the effectiveness of adherence interventions may have a far 7 8 greater impact on the health of the population than any improvement in specific medical 9 10 11 treatments”.[23] 12 13 The data of this study show also that for the same active substance, a change of product 14 15 For peer review only 16 (regardless of whether it is a brand name or generic drug) increases the risk of 17 18 discontinuation of therapy and of substitution with another molecule of the same class. Our 19 20 findings are comparable with those reported by Ghate et al.,[24] who found that switching 21 22 among warfarin formulations, including substituting a generic for another generic, may 23 24 expose patients with atrial fibrillation to a higher risk of thrombotic and bleeding events than 25 26 27 remaining on the same formulation. 28 29 At present, only few studies have estimated the frequency and effects of substitution 30 31 32 between different products of the same active substance in a clinical practice setting. Previous 33
34 analyses suggest that patients switching statin therapy showed significantly poorer http://bmjopen.bmj.com/ 35 36 compliance and higher risk of death or major cardiovascular events when compared with 37 38 controls who did not switch.[25 27] 39 40 41 In addition, our results are also comparable with others that focused on different chronic
42 on September 24, 2021 by guest. Protected copyright. 43 therapies, such as a recent study by Kesselheim et al. showing that changes in pill colors and 44 45 shapes increased the risk of non adherence among epileptic patients.[28] The possibility that 46 47 48 variation in packaging and pill appearance, as well as in the shape and color of either box or 49 50 tablet, may affect adherence is a reason for concern. 51 52 53 Moreover, we cannot exclude the possibility that other potential determinants can play a 54 55 key role in a reduction of patients’ adherence. Observational studies[29, 30] have 56 57 demonstrated a relationship between age, gender, cardiovascular risk factors, polytherapies 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 15 of 23 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 and a suboptimal adherence to therapy; our findings are in agreement with these previous 5 6 analyses. 7 8 9 However, the underlying reasons for poor adherence are not fully understood, and there may 10 11 be many reasons behind these behaviours, some of which relate to the perceptions that 12 13 physicians, pharmacists and patients may have of drugs and therapies. Nevertheless, the 14 15 For peer review only 16 clinical consequences that may result from these perceptions are important and should be 17 18 considered. Establishing better physician patient communication, improving patient 19 20 education, and maintaining regular follow up and review of patients’ progress may be as 21 22 important as other factors in encouraging adherence and lead to improved health outcomes 23 24 and enhanced patient safety.[11] This includes addressing patients’ perceptions about the 25 26 27 medications they are prescribed and understanding that they may find routine changes in the 28 29 name and appearance of long term medications challenging. 30 31 32 There are several limitations to this study. The first, in common with other similar studies, is 33
34 that the reasons for switch, non adherence or discontinuation of treatment are not retrievable http://bmjopen.bmj.com/ 35 36 from the data set. A second limitation is a relatively limited sample size. 37 38 39 Although in our study we used the healthcare databases of Lombardy, Lazio and Campania, 40 41 three Italian Regions localized from north to south of Italy and with a total population of
42 on September 24, 2021 by guest. Protected copyright. 43 about 21 million, and considering that we have focused our analysis among users of 44 45 simvastatin, ramipril, and amlodipine, large studies are thus needed to confirm and to 46 47 48 enhance the generalizability of the findings, and in different populations. 49 50 Despite these limitations, our study indicates that in a “real world” setting, changes among 51 52 53 different product of the same active substance, including switching brand name to generic, 54 55 generic to another generic and generic to brand name, are quite common among patients with 56 57 CVD. Our findings suggest that switching to a different product of the same off patent active 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 16 of 23 BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 substance, brand name or generic, may expose patients to a higher risk of therapy 5 6 discontinuation or substitution than continuing treatment with the same product. 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33
34 http://bmjopen.bmj.com/ 35 36 37 38 39 40 41
42 on September 24, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 17 of 23 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 Acknowledgements: We thank Ray Hill and Gayle Robins, independent medical writers, 5 6 who provided English language editing and journal styling prior to submission. 7 8 9 Contributors: LDE, FS, DS conceived and designed the study; LDE, DS, SB, EDE analyzed 10 11 the data; and LDE, FS, DS wrote the paper. All authors revised and approved the final 12 13 version. 14 15 For peer review only 16 Funding: This research received no specific grant from any funding agency in the public, 17 18 commercial or not for profit sectors. 19 20 Competing interests: All authors have completed the Unified Competing Interest form at 21 22 http://www.icmje.org/coi_disclosure.pdf (available on request from the corresponding author) 23 24 and declare no support from any organization for the submitted work. 25 26 27 FS has served as consultant and had speaking engagements to pharma companies marketing 28 29 cardiovascular drugs and has been compensated for travel and time spent on research and 30 31 lectures. 32 33 Ethics approval: For this type of study formal consent is not required. However, to 34 http://bmjopen.bmj.com/ 35 36 guarantee patient privacy, no personal identifiers were provided to the researchers. According 37 38 to the Italian law for confidentiality of data the study was notified to the Ethic Committees of 39 40 each Local Health Unit. 41
42 Data Sharing: No additional data available on September 24, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 18 of 23 BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 References 5 6 7 1. World Health Organization. The top 10 causes of death. 2014. Available at: 8 9 http://www.who.int/mediacentre/factsheets/fs310/en/index.html (accessed 10 11 September 30, 2014) 12 13 2. Wald NJ, Law MR. A strategy to reduce cardiovascular disease by more than 14 15 80%. BMJFor 2003;326:1419 23. peer review only 16 17 18 3. Naderi SH, Bestwick JP, Wald DS. Adherence to drugs that prevent 19 20 cardiovascular disease: meta analysis on 376,162 patients. Am J Med 21 22 2012;125:882 7.e1 23 24 4. Burke LE, Dunbar Jacobs JM, Hill MN. Compliance with cardiovascular disease 25 26 27 prevention strategies: a review of the research. Ann Behav Med 2012;19:239 63. 28 29 5. Gabbay U, Yosef N, Feder Krengel N, et al. Therapeutic equivalent substitute that is new 30 31 or unfamiliar to the chronic patient may result in medication error. Int J Health Care 32 33 Qual Assur 2012;25:509 18. 34 http://bmjopen.bmj.com/ 35 36 6. Greene JA. The substance of the brand. Lancet 2011;378:120 1. 37 38 7. Posner J, Griffin JP. Generic substitution. Br J Clin Pharmacol 2011;72:731 2. 39 40 8. Greene JA, Kesselheim AS. Why do the same drugs look different? Pills, trade 41
42 dress, and public health. N Engl J Med 2011;365:83 9. on September 24, 2021 by guest. Protected copyright. 43 44 9. Hakonsen H, Eilertsen M, Borge H, et al. Generic substitution: additional 45 46 47 challenge for adherence in hypertensive patients? Curr Med Res Opin 48 49 2009;25:2515 21. 50 51 10. Atar D, Carmena R, Clemmensen P, et al. Clinical review: impact of statin 52 53 substitution policies on patient outcomes. Ann Med 2009;41:242 56. 54 55 56 11. World Health Organization. Adherence to long term therapies: evidence for 57 58 action. 2003.WHO, Geneva, Switzerland. Available from 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 19 of 23 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 http://www.emro.who.int/ncd/Publications/adherence_report.pdf (accessed 5 6 September 30, 2014). 7 8 12. Ude M, Schuessel K, Quinzler R, et al. Generic switch after ramipril patent 9 10 11 expiry is not associated with decreased pharmacy refill compliance: a 12 13 retrospective study using the DAPI database. J Hypertens 2011;29:1837 45. 14 15 13. Van WijkFor BL, Klungel peer OH, Heerdink review ER, et al. Generic only substitution of 16 17 antihypertensive drugs: does it affect adherence? Ann Pharmacother 2006;40:15 18 19 20 20. 21 22 14. Chapman RH, Benner JS, Girase P, et al. Generic and therapeutic statin switches and 23 24 disruptions in therapy. Curr Med Res Opin 2009;25:1247 60. 25 26 15. Colombo GL, Agabiti Rosei E, Margonato A, et al. Off patent generic medicines 27 28 vs. off patent brand medicines for six reference drugs: a retrospective claims 29 30 31 data study from five local healthcare units in the Lombardy Region of Italy. 32 33 PLoS One 2013;8:e82990.
34 http://bmjopen.bmj.com/ 35 16. Hakonsen H, Toverud EL. Special challenges for drug adherence following 36 37 generic substitution in Pakistani immigrants living in Norway. Eur J Clin 38 39 40 Pharmacol 2011;67:193 201. 41
42 17. Degli Esposti L, Sangiorgi D, Buda S, et al. Discontinuità terapeutica nei pazienti on September 24, 2021 by guest. Protected copyright. 43 44 sottoposti a sostituzione tra farmaci equivalenti. Evidenze dal “mondo reale”. 45 46 Supplemento a Politiche sanitarie 2015;16:3. 47 48 18. Agenzia Italiana del Farmaco (AIFA). Guideline for the classification and conduction of 49 50 51 the observational studies on medicines. 2010. Available from: 52 53 https://www.agenziafarmaco.gov.it/ricclin/sites/default/files/files_wysiwyg/files/CIRCU 54 55 LARS/Circular%2031st%20May%202010.pdf (accessed September 30, 2014). 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 20 of 23 BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 19. Steinman MA, Landefeld CS, Rosenthal GE, et al. Polypharmacy and prescribing 5 6 quality in older people. J Am Geriatr Soc 2006;54:1516 23. 7 8 20. Volpe M, Chin D, Paneni F. The challenge of polypharmacy in cardiovascular 9 10 11 medicine. Fundam Clin Pharmacol 2010;24:9 17. 12 13 21. Frankenstein L, Clark AL, Ribeiro JP. Influence of sex on treatment and outcome 14 15 in chronicFor heart failure. peer Cardiovasc review Ther 2012;30:182 92. only 16 17 22. Barsheshet A, Brenyo A, Goldenberg I, et al. Sex related differences in patients' 18 19 20 responses to heart failure therapy. Nat Rev Cardiol 2012;9:234 42. 21 22 23. Haynes RB, McDonald H, Garg AX, et al. Interventions for helping patients to 23 24 follow prescriptions for medications. Cochrane Database Syst Rev 25 26 2002;(2):CD000011. 27 28 24. Ghate SR, Biskupiak JE, Ye X, et al. Hemorrhagic and thrombotic events 29 30 31 associated with generic substitution of warfarin in patients with atrial 32 33 fibrillation: a retrospective analysis. Ann Pharmacother 2011;45:701 12.
34 http://bmjopen.bmj.com/ 35 25. Thiebaud P, Patel BV, Nichol MB, et al. The effect of switching on compliance 36 37 and persistence: the case of statin treatment. Am J Manag Care 2005;11:670 4. 38 39 40 26. Phillips B, Roberts C, Rudolph A, et al. Switching statins: the impact on patient 41
42 outcomes. Br J Cardiol 2007;14:280 5. on September 24, 2021 by guest. Protected copyright. 43 44 27. Liew D, Webb K, Meerding W J, et al. Potential cardiovascular consequences of 45 46 switching from atorvastatin to generic simvastatin in the Netherlands. Neth 47 48 Heart J 2012;20:197 201. 49 50 51 28. Kesselheim AS, Misono AS, Shrank WH, et al. Variations in pill appearance of 52 53 antiepileptic drugs and the risk of nonadherence. JAMA Intern Med 2013;173:202 8. 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 21 of 23 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 29. Perreault S, Blais L, Lamarre D, et al. Persistence and determinants of statin therapy 5 6 among middle aged patients for primary and secondary prevention. Br J Clin Pharmacol 7 8 2005;59:564 573. 9 10 11 30. Shin JY, Choi NK, Jung SY,et al. Overlapping medication associated with healthcare 12 13 switching among Korean elderly diabetic patients. J Korean Med Sci 2011;26:1461 8. 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33
34 http://bmjopen.bmj.com/ 35 36 37 38 39 40 41
42 on September 24, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 22 of 23 BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 Figure legend 5 6 7 Fig. 1 Definitions of the patient cohorts 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33
34 http://bmjopen.bmj.com/ 35 36 37 38 39 40 41
42 on September 24, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 23 of 23 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 Figure 1. Definitions of the patient cohorts 33
34 254x190mm (96 x 96 DPI) http://bmjopen.bmj.com/ 35 36 37 38 39 40 41
42 on September 24, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
Therapy discontinuation or substitution in patients with cardiovascular disease, switching among different products of the same off-patent active substance: a “real-world” retrospective cohort study
For peer review only Journal: BMJ Open
Manuscript ID bmjopen-2016-012003.R1
Article Type: Research
Date Submitted by the Author: 24-May-2016
Complete List of Authors: Degli Esposti, Luca; CliCon S.r.l, Health, Economics and Outcomes Research, Sangiorgi, Diego; CliCon S.r.l. Health, Economics and Outcomes Research Buda, Stefano; CliCon S.r.l. Health, Economics and Outcomes Research Degli Esposti, Ezio; CliCon S.r.l. Health, Economics and Outcomes Research Scaglione, Francesco ; University of Milan, Medical Biotechnology and Translational Medicine
Primary Subject Cardiovascular medicine Heading:
Secondary Subject Heading: Public health http://bmjopen.bmj.com/ Keywords: Simvastatin, Amlodipine, Ramipril, Switching, Adherence to treatment
on September 24, 2021 by guest. Protected copyright.
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 1 of 34 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 5 Therapy discontinuation or substitution in patients with cardiovascular disease, 6 7 switching among different products of the same off-patent active substance: a “real- 8 9 world” retrospective cohort study 10 11 12 13 14 Luca Degli Esposti,1 Diego Sangiorgi,1 Stefano Buda,1 Ezio Degli Esposti,1 Francesco 15 For peer review only 16 2 17 Scaglione 18 19 20 21 22 1CliCon S.r.l. Health, Economics and Outcomes Research, Ravenna, Italy 23 24 2 25 Department of Medical Biotechnology and Translational Medicine, University of Milan, 26 27 Milan, Italy 28 29 30 31 32 Correspondence to 33
34 http://bmjopen.bmj.com/ 35 Luca Degli Esposti, EconD., CliCon S.r.l, Health, Economics and Outcomes Research, Via 36 37 Salara, 36 48100 Ravenna, Italy 38 39 40 Tel +39 544 38393 Fax +39 544 212699. [email protected] 41
42 on September 24, 2021 by guest. Protected copyright. 43 44 45 Keywords 46 47 48 Simvastatin, Amlodipine, Ramipril, Switching, Adherence to treatment 49 50 51 52 53 Word count: 2,452 54 55 56 57 58 59 1 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 2 of 34 BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 ABSTRACT 5 6 Objective: The present study investigated the effects of switching to different products of the 7 8 same off patent active substance (brand name or generic) on therapy discontinuation or 9 10 11 substitution with another molecule of the same class, in patients with cardiovascular disease 12 13 treated with statins and antihypertensives in clinical practice. 14 15 For peer review only 16 Design: A retrospective cohort study in a “real world” setting. 17 18 Setting: Analysis of data performed by integrating administrative databases that included a 19 20 21 total of approximately two million health assisted individuals from three Italian Local Health 22 23 Units located in three different regions of Italy. 24 25 Participants: All patients aged ≥18 years with at least one prescription of simvastatin, 26 27 st st 28 ramipril or amlodipine in the period January 1 to December 31 2010 were included and 29 30 followed up for two years. 31 32 33 Main outcome measures: Prescription refills occurring during follow up were evaluated.
34 http://bmjopen.bmj.com/ 35 Frequency of discontinuation of therapy or substitution with another molecule of the same 36 37 class (for example, from simvastatin to a different statin) during follow up was identified. 38 39 40 Results: During follow up, therapy discontinuation or substitution was found to be more 41
42 frequent in patients switching to a different product of the same active substance compared on September 24, 2021 by guest. Protected copyright. 43 44 with non switching patients (11.5% vs 10.8% and 22.2% vs 20.8% [p=0.002], respectively, 45 46 in the simvastatin group; 7.6% vs 7.8% and 19.2% vs 17.0% [p<0.001], respectively, in the 47 48 49 ramipril group; 4.0% vs 3.5% and 24.6% vs 22.7% [p<0.001], respectively, in the amlodipine 50 51 group). These findings were partially confirmed by multivariate analysis. 52 53 54 Conclusions: Switches among products of the same active substance are quite common in 55 56 patients with cardiovascular disease. Our study suggests that switching may expose patients 57 58 to a higher risk of therapy discontinuation or substitution. 59 2 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 3 of 34 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 5 6 7 STRENGTHS AND LIMITATIONS OF THIS STUDY 8 9 • This study, in a “real world” setting, is one of only a few studies to investigate 10 11 12 clinical differences related to switching among different products of the same active 13 14 substance in the cardiovascular setting. Until now, most research has focused only on 15 For peer review only 16 comparing brand name and generic drugs. 17 18 19 • Sample size was relatively limited, and although we used three healthcare databases 20 21 comprising a total of approximately two million health assisted individuals from three 22 23 regions of Italy, larger studies are needed to confirm and to enhance the 24 25 generalizability of the findings, and in different populations. 26 27 28 • In common with other retrospective, observational studies, reasons for switch, non 29 30 adherence or discontinuation of treatment were not retrievable from the data set. 31 32 33
34 http://bmjopen.bmj.com/ 35 36 37 38 39 40 41
42 on September 24, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 3 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 4 of 34 BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 INTRODUCTION 5 6 7 Cardiovascular diseases (CVD) are the leading cause of death worldwide, accounting for 8 9 approximately one third of all deaths.[1] Combination therapy with antihypertensive drugs 10 11 12 and serum cholesterol lowering drugs is effective in prevention, and it is estimated that a high 13 14 level of adherence to treatment will reduce the risk of CVD by approximately 80%.[2] A 15 For peer review only 16 number of studies have demonstrated that patients often discontinue long term treatment or 17 18 take less than prescribed, and that such non adherence reduces the potential preventive 19 20 benefits.[3] 21 22 23 Many reasons contribute to patient non adherence to medical therapy, such as ageing, 24 25 comorbidities, poor relationship between patient and physician, poor memory, and patients’ 26 27 28 low perception of disease severity.[4] In addition, the need to take several drugs 29 30 concomitantly, or other medication related factors, may make remembering when to take 31 32 each drug more difficult and increase the risk of possible side effects caused by adverse drug 33 34 drug interactions. Although medication side effects are probably not the main cause of poor http://bmjopen.bmj.com/ 35 36 adherence, as there seems to be little direct relationship between adherence and drug class,[3] 37 38 39 they are also associated with treatment discontinuation, especially in the early treatment of 40 41 hypertension [5 8]. Furthermore, some studies have demonstrated that switching between
42 on September 24, 2021 by guest. Protected copyright. 43 different products of the same active substance can have an impact on adherence to 44 45 medication, because variation in packaging and pill appearance may reduce adherence, 46 47 48 especially for chronic diseases.[9, 10] 49 50 There is a perception among patients and physicians alike that frequent changes between 51 52 53 branded and unbranded products (as well as between generics), all containing the same active 54 55 substance, and especially if patients are older and on multi drug regimens, may cause patients 56 57 to become anxious when the appearance of their drugs changes.[11 13] This can lead to an 58 59 4 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 5 of 34 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 increased risk of patients making mistakes or double medicating, which flows on to increased 5 6 drug non adherence.[14, 15] 7 8 9 Few studies have investigated clinical differences related to switching among different 10 11 products of the same active substance in the cardiovascular setting. Until now, most research 12 13 has focused only on comparing brand name and generic drugs.[16 20] 14 15 For peer review only 16 The aim of the present study was to investigate the effects of switching to different 17 18 products of the same off patent active substance (brand name or generic) on therapy 19 20 21 discontinuation or substitution with another molecule of the same class, in patients with CVD 22 23 treated with statins and antihypertensives in clinical practice. 24 25 A version of this article has been previously published as a journal supplement in the 26 27 28 Italian language.[21] 29 30 31 METHODS 32 33 Data collection 34 http://bmjopen.bmj.com/ 35 36 The data used for the analysis were obtained from the administrative databases of three Local 37 38 Health Units (LHUs), whose databases included a total population of about two million 39 40 41 health assisted individuals, in the Italian regions of Lombardy, Lazio and Campania. We
42 on September 24, 2021 by guest. Protected copyright. 43 analyzed the following archives: Assisted Subjects’ Database, containing the personal data of 44 45 patients; Medication Prescription database, containing all the information relating to 46 47 individual prescriptions, such as the International Nonproprietary Names (INN) for 48 49 50 pharmaceutical substances, the Anatomical Therapeutic Chemical (ATC) code of the 51 52 prescribed drug, the number of packages, the number of units per package, the dose, the cost 53 54 per unit and the date of the prescription; Hospital Discharge Database (SDO), containing 55 56 information on each hospital discharge, in particular the date of admission and discharge, 57 58 59 5 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 6 of 34 BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 primary and secondary diagnoses coded according to the International Classification of 5 6 Diseases, Ninth Revision, Clinical Modification (ICD-9-CM). The patient code in each 7 8 database allowed electronic linking among all databases. To guarantee patient privacy, this 9 10 11 patient code was transcoded into an anonymous univocal numeric code. No identifiers related 12 13 to patients were provided to the researchers. According to the Italian law for confidentiality 14 15 of data[22] theFor study was peer notified to the Ethicreview Committees of eachonly LHU. 16 17 18 Cohort definition 19 20 21 The study was a retrospective cohort study including all patients aged ≥18 years that,
22 st st 23 between January 1 2010 and December 31 2010 (enrollment period), had at least one 24 25 prescription of simvastatin [ATC code: C10AA01], ramipril [ATC code: C09AA05] or 26 27 amlodipine [ATC code: C08CA01]. The date of enrollment was defined as the earliest date 28 29 within the enrollment period in which the patient had the last switch of medication 30 31 32 (switchers) or the last prescription in the case of a patient continuing with the same 33
34 medication (non switchers). Starting from this date, the individual patient was followed for http://bmjopen.bmj.com/ 35 36 two years (follow up period). Switchers were defined as those patients who switched among 37 38 different products of the same off patent active substance (i.e., from brand name to generic, 39 40 41 from generic to brand name, or from generic to another generic). The changes in dose and
42 on September 24, 2021 by guest. Protected copyright. 43 dosage form were not accounted for during switching. The definitions of the patient cohorts 44 45 are shown in Figure 1. Data on baseline characteristics, including demographics, 46 47 cardiovascular risk factors and polymedication, were collected; specifically, polymedication 48 49 were identified as more than one cardiovascular medication at the date of enrollment. The 50 51 52 data on drug prescriptions and hospitalizations that occurred during the 12 months preceding 53 54 the date of enrollment were analyzed (characterization period). 55 56 57 58 59 6 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 7 of 34 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 Only patients with at least one prescription of the index drug in the previous 12 months 5 6 were included (to capture patients who could have made a change in therapy) and with at 7 8 least two prescriptions at follow up (to include patients with continuity of treatment). Patients 9 10 11 treated with fixed combinations of the molecules under consideration (ramipril and diuretics 12 13 [ATC code: C09BA05], perindopril and amlodipine [ATC code: C09BB04], ramipril and 14 15 felodipine [ATCFor code: C09BB05],peer olmesartan review and amlodipine [ATConly code: C09DB02], 16 17 simvastatin and ezetimibe [ATC code: C10BA02]) were excluded. 18 19 20 Patients transferred to another LHU during the follow up period were excluded from the 21 22 analysis. 23 24 25 Study population 26 27 28 Cardiovascular risk 29 30 Patients were classified as high cardiovascular risk if they had cardiovascular treatment or 31 32 33 hospitalization for diabetes. Hospitalizations related to diabetes were identified by ICD 9 CM
34 http://bmjopen.bmj.com/ 35 code: 250 (primary discharge reasons); and/or cardiovascular risk factors (previous 36 37 hospitalization for Ischemic Heart Disease [Acute Myocardial Infarction (ICD 9 CM: 410) 38 39 Acute Cardiac Ischemia (ICD 9 CM: 411), Old Myocardial Infarction (ICD 9 CM: 412), 40 41 Angina Pectoris (ICD 9 CM: 413), Chronic Cardiac Ischemia (ICD 9 CM: 414)]; Cerebral
42 on September 24, 2021 by guest. Protected copyright. 43 44 Hemorrhage [ICD 9 CM: 431]; Cerebral Artery Occlusion [ICD 9 CM: 434]; Transient 45 46 Cerebral Ischemia [ICD 9 CM: 435]; Cerebral Circulatory Disorders [ICD 9 CM: 436]; 47 48 Atherosclerosis [ICD 9 CM: 440]; Other Peripheral Vascular Disease [ICD 9 CM: 443]; 49 50 Chronic Renal Failure [ICD 9 CM: 585]; Coronary Angioplasty [ICD 9 CM procedure: 51 52 53 0066, 360]); and/or at least two prescriptions of antidiabetic drugs [ATC code: A10]. All 54 55 other patients were classified as at moderate cardiovascular risk. 56 57 58 Drug treatments 59 7 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 8 of 34 BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 Patients were also characterized by strategy of treatment at baseline with lipid lowering drugs 5 6 (ATC code: C10) and antihypertensive drugs (ATC codes: C02, C03, C07, C08, C09). 7 8 9 Data analysis at follow up 10 11 12 During the follow up period, the discontinuation or the first substitution of therapy was 13 14 identified. Discontinuation of therapy was defined as the absence of prescriptions of the same 15 For peer review only 16 therapeutic class (ATC group) as the index molecule in the last quarter of observation. A 17 18 substitution of therapy was defined as a change to a different active substance of the same 19 20 21 therapeutic class (ATC group) (i.e., switching from simvastatin to a different statin). A switch 22 23 among different products of the same active substance was identified by INN. 24 25 Statistical analysis 26 27 28 Continuous variables were reported as mean ± standard deviation (SD) and compared using 29 30 the Student's t test; categorical variables were reported as absolute numbers and percentages 31 32 33 and compared using the chi square test.
34 http://bmjopen.bmj.com/ 35 Discontinuations of therapy and substitution with another molecule of the same class 36 37 38 were analyzed by multivariate analysis using Cox proportional hazards models; covariates 39 40 considered in the models were: age, male sex, high cardiovascular risk, cardiovascular 41
42 treatments, change of formulation in the period of characterization. on September 24, 2021 by guest. Protected copyright. 43 44 45 The analysis of Schoenfeld residuals (scaled and unscaled) was conducted to assess the 46 47 proportionality of risk. 48 49 50 p Values <0.05 were considered statistically significant. All analyses were performed 51 52 using STATA 12.0 SE. 53 54 55 RESULTS 56 57 58 Simvastatin 59 8 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 9 of 34 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 A total of 38,183 patients treated with simvastatin, 17,642 male (46%), mean age 68.3±10.7 5 6 years, were included in the analysis. A total of 9,392 (25%) patients were classified as at high 7 8 cardiovascular risk, while 30,467 (80%) received concomitant cardiovascular treatments 9 10 11 (Table 1). 12 13 Table 1 Characteristics of the enrolled patients 14 15 For peer review only 16 Simvastatin Ramipril Amlodipine 17 18 Non- p Non- p Non- p 19 Category switchers Switchers Value switchers Switchers Value switchers Switchers Value 20 21 22 Total 23,180 (61) 15,003 (39) 22,799 (71) 9,312 (29) 26,823 (72) 10,644 (28) 23 24 Patient Characteristics 25 26 Age, y 68.5 ± 10.7 68.0 ± 10.6 <0.001 67.0 ± 12.7 66.8 ± 13.0 0.203 68.3 ± 11.7 68.1 ± 11.8 0.137 27 28 29 Male sex 10,469 (45.2) 7,173 (47.8) <0.001 12,876 5,617 (60.3) <0.001 14,126 6,213 (58.4) <0.001 30 (56.5) (52.7) 31 32 High CVR 5,494 (23.7) 3,898 (26.0) <0.001 4,873 (21.4) 2,025 (21.7) 0.470 4,983 (18.6) 2,143 (20.1) <0.001 33
34 http://bmjopen.bmj.com/ 35 Additional 18,539 (80.0) 11,928 0.265 17,868 7,393 (79.4) 0.044 23,793 9,588 (90.1) <0.001 36 treatments (79.5) (78.4) (88.7) 37 38 39 Values are N (%) or mean ± standard deviation (SD). 40 41 CVR, cardiovascular risk.
42 on September 24, 2021 by guest. Protected copyright. 43 44 Switches among different products occurred in 39% of patients treated with simvastatin. 45 46 Switcher patients were mainly men with high cardiovascular risk; this cohort was slightly 47 48 younger than non switchers, the difference was statistically significant. With regard to 49 50 51 switchers, a little over half carried out one switch only during the characterization period, 52 53 with 8% having four switches or more (Table 2). 54 55 56 57 58 59 9 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 10 of 34 BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 Table 2 Annual frequency of switches 5 6 7 Simvastatin Ramipril Amlodipine 8 9 10 Switches (N) Patients (%) Switches (N) Patients (%) Switches (N) Patients (%) 11 12 1 7,842 (52) 1 5,112 (55) 1 5,710 (54) 13 14 15 2 For4,062 peer (27) 2review 2,501 (27) only2 2,871 (27) 16 17 3 1,917 (13) 3 1,116 (12) 3 1,300 (12) 18 19 4 805 (5) 4 413 (4) 4 563 (5) 20 21 22 ≥5 377 (3) ≥5 170 (2) ≥5 200 (2) 23 24 25 26 27 In the follow up period, 4,232 (11%) patients undertook a therapy substitution with another 28 29 molecule; a significantly lower percentage of substitution was observed in the group that did 30 31 not switch to a different product of the same active substance (Table 3). 32 33
34 http://bmjopen.bmj.com/ 35 36 37 38 39 40 41
42 on September 24, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 10 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 11 of 34 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 Table 3 Therapy discontinuation and substitution in switcher and non switcher patients 5 6 7 Simvastatin Ramipril Amlodipine 8 9 Switch Switch Switch 10 11 No Yes p No Yes p No Yes p 12 13 Value Value Value 14 15 Total, N (%) For23,180 peer15,003 review 22,799 (71) 9,312 (29) only 26,823 (72) 10,644 (28) 16 17 (61) (39) 18 19 Therapy substitution or discontinuation 20 21 Substitution, N (%) 2,506 1,726 0.037 1,785 (7.8) 711 (7.6) 0.571 943 (3.5) 426 (4.0) 0.026 22 23 (10.8) (11.5) 24 25 Discontinuation, N (%) 4,826 3,327 0.002 3,885 (17.0) 1,792 (19.2) <0.001 6,089 (22.7) 2,618 (24.6) <0.001 26 (20.8) (22.2) 27 28 29 30 31 32 In the same period 8,153 (21%) patients discontinued treatment; a significantly lower 33
34 percentage of discontinuation was observed for non switching patients (Table 3). These http://bmjopen.bmj.com/ 35 36 findings were partially confirmed by multivariate analysis (Table 4): the group that switched 37 38 to a different product of the same active substance showed a higher probability of 39 40 41 discontinuation (hazard ratio [HR]=1.087, p<0.001) and a higher, but not significant,
42 on September 24, 2021 by guest. Protected copyright. 43 probability of substitution of therapy (HR=1.059, p=0.068). 44 45 Table 4 Multivariate analysis of predictors of risk of therapy discontinuation and substitution 46 47 48 49 Variable Simvastatin 50 51 Substitution Discontinuation 52 53 HR 95% CI p Value HR 95% CI p Value 54 55 56 Age 0.993 0.990 0.995 <0.001 1.001 0.999 1.004 0.194 57 58 59 11 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 12 of 34 BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 Male sex 1.116 1.050 1.186 <0.001 0.921 0.882 0.963 <0.001 5 6 7 High CVR 1.107 1.034 1.185 0.004 0.962 0.914 1.013 0.140 8 9 Additional CV treatments 1.064 0.984 1.152 0.121 0.821 0.778 0.867 <0.001 10 11 Switch to different product of 12 13 the same substance 1.059 0.996 1.126 0.068 1.087 1.040 1.136 <0.001 14 15 For peer reviewRamipril only 16 17 18 Substitution Discontinuation 19 20 HR 95% CI p Value HR 95% CI p Value 21 22 Age 1.003 1.000 1.006 0.072 1.005 1.002 1.007 <0.001 23 24 25 Male sex 0.938 0.865 1.017 0.119 0.950 0.900 1.002 0.058 26 27 High CVR 1.158 1.057 1.269 0.002 0.866 0.811 0.926 <0.001 28 29 Additional CV treatments 1.521 1.360 1.700 <0.001 0.806 0.757 0.857 <0.001 30 31 32 Switch to different product of 33 the same substance 0.973 0.892 1.062 0.540 1.163 1.100 1.230 <0.001
34 http://bmjopen.bmj.com/ 35 36 Amlodipine 37 38 Substitution Discontinuation 39 40 HR 95% CI p Value HR 95% CI p Value 41
42 on September 24, 2021 by guest. Protected copyright. 43 Age 1.019 1.013 1.024 <0.001 1.006 1.004 1.008 <0.001 44 45 Male sex 0.807 0.719 0.907 <0.001 0.870 0.834 0.909 <0.001 46 47 High CVR 1.247 1.091 1.424 0.001 0.879 0.831 0.929 <0.001 48 49 50 Additional CV treatments 1.935 1.514 2.474 <0.001 0.959 0.897 1.025 0.215 51 52 Switch to different product of 53 54 the same substance 1.179 1.043 1.333 0.008 1.124 1.074 1.177 <0.001 55 56 CI, confidence interval; CV, cardiovascular; CVR, cardiovascular risk; HR, hazard ratio. 57 58 59 12 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 13 of 34 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 Ramipril 5 6 7 A total of 32,111 patients treated with ramipril, 18,493 male (58%), mean age 66.9±12.8 8 9 years, were included in the analysis. Of these, 6,898 (21%) patients were classified as at high 10 11 cardiovascular risk, while 25,261 (79%) were receiving additional cardiovascular treatments 12 13 (Table 1). Switches among different products occurred in 29% of patients treated with 14 15 For peer review only 16 ramipril. Switcher patients were mainly men and the difference was statistically significant. 17 18 With regard to switchers, again, a little over half (55%) carried out one switch only, during 19 20 the characterization period, and few (6%) had four switches or more (Table 2). In the follow 21 22 up period, 2,496 (8%) patients undertook a therapy substitution to another molecule; no trend 23 24 towards a lower percentage of substitution was observed in the group that did not switch to a 25 26 27 different product of the same active substance (Table 3). In the same period 5,677 (18%) 28 29 patients discontinued treatment, with a significant difference in favor of patients not 30 31 switching to a different product of the same active substance (Table 3). These findings were 32 33 confirmed by multivariate analysis (Table 4): there was essentially no difference between 34 http://bmjopen.bmj.com/ 35 36 groups in terms of probability of substitution (HR=0.973, p=0.540), while the non switching 37 38 group showed a significantly lower probability of discontinuation of therapy (HR=1.163, 39 40 p<0.001). 41
42 on September 24, 2021 by guest. Protected copyright. 43 Amlodipine 44 45 A total of 37,467 patients treated with amlodipine, 20,339 male (54%), mean age 68.2±11.7 46 47 48 years, were included in the analysis. Of these, 7,126 (19%) patients were classified as at high 49 50 cardiovascular risk, while 33,381 (89%) were receiving additional cardiovascular treatments 51 52 (Table 1). Switches among different products occurred in 28% of patients treated with 53 54 amlodipine. Switcher patients were mainly men with high cardiovascular risk; the difference 55 56 57 was statistically significant. With regard to switchers, just over half (54%) carried out one 58 59 13 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 14 of 34 BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 switch only during the characterization period, and 7% had four switches or more (Table 2). 5 6 In the follow up period, 1,369 (4%) patients undertook a therapy substitution to another 7 8 molecule; a significantly lower percentage of substitution was observed in the group that did 9 10 11 not switch to a different product of the same active substance (Table 3). In the same period 12 13 8,707 (23%) patients discontinued treatment; a significantly lower probability of 14 15 discontinuationFor was observed peer for patients review not switching to another only product of the same active 16 17 substance (Table 3). These findings were confirmed by multivariate analysis (Table 4): the 18 19 20 switcher group showed a higher probability of discontinuation (HR=1.124, p<0.001) and 21 22 substitution of therapy (HR=1.179, p=0.008). 23 24 25 DISCUSSION 26 27 28 In accordance with previous studies,[23 26] this retrospective analysis in a “real world” 29 30 setting shows that age, gender, cardiovascular risk, and polymedication affect discontinuation 31 32 of therapy. A number of factors may interact to affect adherence to therapies for chronic 33 34 conditions. These have been categorized by the World Health Organization as social and http://bmjopen.bmj.com/ 35 36 economic related factors, health system/health care team related factors, condition related 37 38 39 factors, and patient related factors.[15] As poor adherence has a significant negative impact 40 41 on health outcomes and healthcare costs, and imposes a substantial burden on patients and
42 on September 24, 2021 by guest. Protected copyright. 43 health systems, “increasing the effectiveness of adherence interventions may have a far 44 45 greater impact on the health of the population than any improvement in specific medical 46 47 48 treatments”.[27] 49 50 The data of this study show also that for the same active substance, a change of product 51 52 53 (regardless of whether it is a brand name or generic drug) increases the risk of 54 55 discontinuation of therapy and of substitution with another molecule of the same class. Our 56 57 findings are comparable with those reported by Ghate et al.,[28] who found that switching 58 59 14 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 15 of 34 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 among warfarin formulations, including substituting a generic for another generic, may 5 6 expose patients with atrial fibrillation to a higher risk of thrombotic and bleeding events than 7 8 remaining on the same formulation. 9 10 11 At present, only few studies have estimated the frequency and effects of substitution 12 13 between different products of the same active substance in a clinical practice setting. Previous 14 15 For peer review only 16 analyses suggest that patients switching statin therapy showed significantly poorer 17 18 compliance and higher risk of death or major cardiovascular events when compared with 19 20 controls who did not switch.[29 31] 21 22 23 In addition, our results are also comparable with others that focused on different chronic 24 25 therapies, such as a recent study by Kesselheim et al. showing that changes in pill colors and 26 27 shapes increased the risk of non adherence among epileptic patients.[32] The possibility that 28 29 variation in packaging and pill appearance, as well as in the shape and color of either box or 30 31 32 tablet, may affect adherence is a reason for concern. 33 34 Moreover, we cannot exclude the possibility that other potential determinants can play a http://bmjopen.bmj.com/ 35 36 37 key role in a reduction of patients’ adherence. Observational studies [33, 34] have 38 39 demonstrated a relationship between age, gender, cardiovascular risk factors, polymedication 40 41 and a suboptimal adherence to therapy; our findings are in agreement with these previous
42 on September 24, 2021 by guest. Protected copyright. 43 analyses. In contrast, new evidence suggested that older age is not related to poorer 44 45 medication adherence to cardiovascular medication. A recent systematic search of the 46 47 48 bibliographic database MedLine and all Cochrane databases, analyzing the relationship 49 50 between age and medication adherence in adult patients with chronic heart failure (CHF), 51 52 showed that older age alone is not related to poorer medication adherence compared with 53 54 younger patients with CHF.[35] Our study does not support this concern. 55 56 57 58 59 15 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 16 of 34 BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 However, the underlying reasons for poor adherence are not fully understood, and there may 5 6 be many reasons behind these behaviours, some of which relate to the perceptions that 7 8 physicians, pharmacists and patients may have of drugs and therapies. Nevertheless, the 9 10 11 clinical consequences that may result from these perceptions are important and should be 12 13 considered. Establishing better physician patient communication, improving patient 14 15 education, andFor maintaining peer regular follow up review and review of only patients’ progress may be as 16 17 important as other factors in encouraging adherence and lead to improved health outcomes 18 19 20 and enhanced patient safety. [15] This includes addressing patients’ perceptions about the 21 22 medications they are prescribed and understanding that they may find routine changes in the 23 24 name and appearance of long term medications challenging. 25 26 27 Our analysis has several limitations inherent to any observational study. First, the study was 28 29 performed using the administrative databases, and the reasons for switch, non adherence or 30 31 discontinuation of treatment in the patients were not retrievable from the dataset. Also, no 32 33 information on the role of the counseling pharmacist when substituting and dispensing drug 34 http://bmjopen.bmj.com/ 35 36 packages were available to us. A second limitation is a relatively limited sample size. 37 38 Although in our study we used the healthcare databases of Lombardy, Lazio and Campania, 39 40 three Italian Regions localized from north to south of Italy, including data for a total 41
42 population of about 2 million, and considering that we have focused our analysis among users on September 24, 2021 by guest. Protected copyright. 43 44 45 of simvastatin, ramipril, and amlodipine, larger studies are needed to confirm and to enhance 46 47 the generalizability of the findings, and in different populations. 48 49 Despite these limitations, our study indicates that in a “real world” setting, changes among 50 51 52 different product of the same active substance, including switching brand name to generic, 53 54 generic to another generic and generic to brand name, are quite common among patients with 55 56 CVD. Our findings suggest that switching to a different product of the same off patent active 57 58 59 16 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 17 of 34 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 substance, brand name or generic, may expose patients to a higher risk of therapy 5 6 discontinuation or substitution than continuing treatment with the same product. 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33
34 http://bmjopen.bmj.com/ 35 36 37 38 39 40 41
42 on September 24, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 17 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 18 of 34 BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 Acknowledgements: We thank Ray Hill and Gayle Robins, independent medical writers, 5 6 who provided English language editing and journal styling prior to submission. 7 8 9 Contributors: LDE, FS, DS conceived and designed the study; LDE, DS, SB, EDE analyzed 10 11 the data; and LDE, FS, DS wrote the paper. All authors revised and approved the final 12 13 version. 14 15 For peer review only 16 Funding: This research received no specific grant from any funding agency in the public, 17 18 commercial or not for profit sectors. 19 20 Competing interests: All authors have completed the Unified Competing Interest form at 21 22 http://www.icmje.org/coi_disclosure.pdf (available on request from the corresponding author) 23 24 and declare no support from any organization for the submitted work. 25 26 27 FS has served as consultant and had speaking engagements to pharma companies marketing 28 29 cardiovascular drugs and has been compensated for travel and time spent on research and 30 31 lectures. 32 33 Data sharing statement: No additional data are available. 34 http://bmjopen.bmj.com/ 35 36 Ethics approval: For this type of study formal consent is not required. However, to 37 38 guarantee patient privacy, no personal identifiers were provided to the researchers. According 39 40 to the Italian law for confidentiality of data the study was notified to the Ethic Committees of 41
42 each Local Health Unit. on September 24, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 18 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 19 of 34 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 References 5 6 7 1. World Health Organization. The top 10 causes of death. 2014. Available at: 8 9 http://www.who.int/mediacentre/factsheets/fs310/en/index.html (accessed 10 11 September 30, 2014) 12 13 2. Wald NJ, Law MR. A strategy to reduce cardiovascular disease by more than 14 15 80%. BMJFor 2003;326:1419 23. peer review only 16 17 18 3. Naderi SH, Bestwick JP, Wald DS. Adherence to drugs that prevent 19 20 cardiovascular disease: meta analysis on 376,162 patients. Am J Med 21 22 2012;125:882 7.e1 23 24 4. Burke LE, Dunbar Jacobs JM, Hill MN. Compliance with cardiovascular disease 25 26 27 prevention strategies: a review of the research. Ann Behav Med 2012;19:239 63. 28 29 5. Kronish IM, Woodward M, Sergie Z, et al. Meta analysis: impact of drug class on 30 31 adherence to antihypertensives. Circulation 2011;123:1611 21. 32 33
34 6. Burke TA, Sturkenboom MC, Lu SE, et al. Discontinuation of antihypertensive http://bmjopen.bmj.com/ 35 36 drugs among newly diagnosed hypertensive patients in UK general practice. J 37 38 Hypertens 2006;24:1193 200. 39 40 41 7. Erkens JA, Panneman MM, Klungel OH, et al. Differences in antihypertensive
42 on September 24, 2021 by guest. Protected copyright. 43 drug persistence associated with drug class and gender: a PHARMO study. 44 45 Pharmacoepidemiol Drug Saf 2005;14:795 803. 46 47 48 8. Friedman O, McAlister FA, Yun L, et al.; Canadian Hypertension Education 49 50 Program Outcomes Research Taskforce. Antihypertensive drug persistence and 51 52 53 compliance among newly treated elderly hypertensives in ontario. Am J Med 54 55 2010;123:173 81. 56 57 58 59 19 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 20 of 34 BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 9. Gabbay U, Yosef N, Feder Krengel N, et al. Therapeutic equivalent substitute that is new 5 6 or unfamiliar to the chronic patient may result in medication error. Int J Health Care 7 8 Qual Assur 2012;25:509 18. 9 10 11 10. Greene JA. The substance of the brand. Lancet 2011;378:120 1. 12 13 11. Posner J, Griffin JP. Generic substitution. Br J Clin Pharmacol 2011;72:731 2. 14 15 12. Greene ForJA, Kesselheim peer AS. Why reviewdo the same drugs lookonly different? Pills, trade 16 17 dress, and public health. N Engl J Med 2011;365:83 9. 18 19 20 13. Hakonsen H, Eilertsen M, Borge H, et al. Generic substitution: additional 21 22 challenge for adherence in hypertensive patients? Curr Med Res Opin 23 24 2009;25:2515 21. 25 26 14. Atar D, Carmena R, Clemmensen P, et al. Clinical review: impact of statin 27 28 substitution policies on patient outcomes. Ann Med 2009;41:242 56. 29 30 31 15. World Health Organization. Adherence to long term therapies: evidence for 32 33 action. 2003.WHO, Geneva, Switzerland. Available from
34 http://bmjopen.bmj.com/ 35 http://www.emro.who.int/ncd/Publications/adherence_report.pdf (accessed 36 37 September 30, 2014). 38 39 40 16. Ude M, Schuessel K, Quinzler R, et al. Generic switch after ramipril patent 41
42 expiry is not associated with decreased pharmacy refill compliance: a on September 24, 2021 by guest. Protected copyright. 43 44 retrospective study using the DAPI database. J Hypertens 2011;29:1837 45. 45 46 17. Van Wijk BL, Klungel OH, Heerdink ER, et al. Generic substitution of 47 48 antihypertensive drugs: does it affect adherence? Ann Pharmacother 2006;40:15 49 50 51 20. 52 53 18. Chapman RH, Benner JS, Girase P, et al. Generic and therapeutic statin switches and 54 55 disruptions in therapy. Curr Med Res Opin 2009;25:1247 60. 56 57 58 59 20 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 21 of 34 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 19. Colombo GL, Agabiti Rosei E, Margonato A, et al. Off patent generic medicines 5 6 vs. off patent brand medicines for six reference drugs: a retrospective claims 7 8 data study from five local healthcare units in the Lombardy Region of Italy. 9 10 11 PLoS One 2013;8:e82990. 12 13 20. Hakonsen H, Toverud EL. Special challenges for drug adherence following 14 15 generic Forsubstitution peer in Pakistani immigrantsreview living in onlyNorway. Eur J Clin 16 17 Pharmacol 2011;67:193 201. 18 19 20 21. Degli Esposti L, Sangiorgi D, Buda S, et al. Discontinuità terapeutica nei pazienti 21 22 sottoposti a sostituzione tra farmaci equivalenti. Evidenze dal “mondo reale”. 23 24 Supplemento a Politiche sanitarie 2015;16:3. 25 26 22. Agenzia Italiana del Farmaco (AIFA). Guideline for the classification and conduction of 27 28 the observational studies on medicines. 2010. Available from: 29 30 31 https://www.agenziafarmaco.gov.it/ricclin/sites/default/files/files_wysiwyg/files/CIRCU 32 33 LARS/Circular%2031st%20May%202010.pdf (accessed September 30, 2014).
34 http://bmjopen.bmj.com/ 35 23. Steinman MA, Landefeld CS, Rosenthal GE, et al. Polypharmacy and prescribing 36 37 quality in older people. J Am Geriatr Soc 2006;54:1516 23. 38 39 40 24. Volpe M, Chin D, Paneni F. The challenge of polypharmacy in cardiovascular 41
42 medicine. Fundam Clin Pharmacol 2010;24:9 17. on September 24, 2021 by guest. Protected copyright. 43 44 25. Frankenstein L, Clark AL, Ribeiro JP. Influence of sex on treatment and outcome 45 46 in chronic heart failure. Cardiovasc Ther 2012;30:182 92. 47 48 26. Barsheshet A, Brenyo A, Goldenberg I, et al. Sex related differences in patients' 49 50 51 responses to heart failure therapy. Nat Rev Cardiol 2012;9:234 42. 52 53 27. Haynes RB, McDonald H, Garg AX, et al. Interventions for helping patients to 54 55 follow prescriptions for medications. Cochrane Database Syst Rev 56 57 2002;(2):CD000011. 58 59 21 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 22 of 34 BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 28. Ghate SR, Biskupiak JE, Ye X, et al. Hemorrhagic and thrombotic events 5 6 associated with generic substitution of warfarin in patients with atrial 7 8 fibrillation: a retrospective analysis. Ann Pharmacother 2011;45:701 12. 9 10 11 29. Thiebaud P, Patel BV, Nichol MB, et al. The effect of switching on compliance 12 13 and persistence: the case of statin treatment. Am J Manag Care 2005;11:670 4. 14 15 30. PhillipsFor B, Roberts peer C, Rudolph A,review et al. Switching statins: only the impact on patient 16 17 outcomes. Br J Cardiol 2007;14:280 5. 18 19 20 31. Liew D, Webb K, Meerding W J, et al. Potential cardiovascular consequences of 21 22 switching from atorvastatin to generic simvastatin in the Netherlands. Neth 23 24 Heart J 2012;20:197 201. 25 26 32. Kesselheim AS, Misono AS, Shrank WH, et al. Variations in pill appearance of 27 28 antiepileptic drugs and the risk of nonadherence. JAMA Intern Med 2013;173:202 8. 29 30 31 33. Perreault S, Blais L, Lamarre D, et al. Persistence and determinants of statin therapy 32 33 among middle aged patients for primary and secondary prevention. Br J Clin Pharmacol
34 http://bmjopen.bmj.com/ 35 2005;59:564 573. 36 37 34. Shin JY, Choi NK, Jung SY,et al. Overlapping medication associated with healthcare 38 39 40 switching among Korean elderly diabetic patients. J Korean Med Sci 2011;26:1461 8. 41
42 35. Krueger K, Botermann L, Schorr SG, et al. Age related medication adherence in patients on September 24, 2021 by guest. Protected copyright. 43 44 with chronic heart failure: A systematic literature review. Int J Cardiol 2015;184:728 35. 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 22 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 23 of 34 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 Figure legend 5 6 7 Fig. 1 Definitions of the patient cohorts 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33
34 http://bmjopen.bmj.com/ 35 36 37 38 39 40 41
42 on September 24, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 23 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 24 of 34 BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 152x69mm (300 x 300 DPI) 26 27 28 29 30 31 32 33
34 http://bmjopen.bmj.com/ 35 36 37 38 39 40 41
42 on September 24, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 25 of 34 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 STROBE Statement—Checklist of items that should be included in reports of cohort studies 4 5 Item No Page 6 Recommendation no. 7 8 Title and abstract 1 (a) Indicate the study’s design with a commonly used term in the title or the 9 abstract 10 11 The title on page 1 in the main manuscript states the study design: 12 1 13 “Therapy discontinuation or substitution in patients with cardiovascular 14 disease, switching among different products of the same off-patent active 15 Forsubstance: peer a “real-world” review retrospective cohort study”only 16 17 (b) Provide in the abstract an informative and balanced summary of what 18 was done and what was found 19 20 The Abstract in the main manuscript provides a balanced summary of what 21 was done and what was found: 2 22 23 ABSTRACT 24 25 Objective: The present study investigated the effects of switching to different 26 products of the same off-patent active substance (brand-name or generic) on 27 therapy discontinuation or substitution with another molecule of the same 28 class, in patients with cardiovascular disease treated with statins and 29 antihypertensives in clinical practice. 30 31 Design: A retrospective cohort study in a “real-world” setting. 32 Setting: Analysis of data performed by integrating administrative databases 33 that included a total of approximately two million health-assisted individuals 34 http://bmjopen.bmj.com/ from three Italian Local Health Units located in three different regions of 35 Italy. 36 37 Participants: All patients aged ≥18 years with at least one prescription of 38 simvastatin, ramipril or amlodipine in the period January 1st to December 39 31st 2010 were included and followed-up for two years. 40 41 Main outcome measures: Prescription refills occurring during follow-up were
42 evaluated. Frequency of discontinuation of therapy or substitution with on September 24, 2021 by guest. Protected copyright. 43 another molecule of the same class (for example, from simvastatin to a 44 different statin) during follow-up was identified. 45 46 Results: During follow-up, therapy discontinuation or substitution was found 47 to be more frequent in patients switching to a different product of the same 48 active substance compared with non-switching patients (11.5% vs 10.8% and 49 22.2% vs 20.8% [p=0.002], respectively, in the simvastatin group; 7.6% vs 50 7.8% and 19.2% vs 17.0% [p<0.001], respectively, in the ramipril group; 4.0% 51 vs 3.5% and 24.6% vs 22.7% [p<0.001], respectively, in the amlodipine 52 group). These findings were partially confirmed by multivariate analysis. 53 Conclusions: Switches among products of the same active substance are 54 quite common in patients with cardiovascular disease. Our study suggests 55 that switching may expose patients to a higher risk of therapy 56 discontinuation or substitution. 57 58 59 60 1 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 26 of 34 BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 Introduction 4 5 Background/ 2 Explain the scientific background and rationale for the investigation being 6 reported 7 rationale 8 In the introduction in the main manuscript a scientific background and 9 rationale is provided: 4,5 10 11 A number of studies have demonstrated that patients often discontinue 12 long-term treatment or take less than prescribed, and that such non- 13 adherence reduces the potential preventive benefits.[3] 14 15 For peerMany reasons contribute review to patient non-adherence only to medical therapy, 16 such as ageing, comorbidities, poor relationship between patient and 17 physician, poor memory, and patients’ low perception of disease severity.[4] 18 In addition, the need to take several drugs concomitantly, or other 19 medication-related factors, may make remembering when to take each drug 20 more difficult and increase the risk of possible side effects caused by adverse 21 drug-drug interactions. Although medication side effects are probably not 22 the main cause of poor adherence, as there seems to be little direct 23 relationship between adherence and drug class,[3] they are also associated 24 with treatment discontinuation, especially in the early treatment of 25 hypertension [5-8]. Furthermore, some studies have demonstrated that 26 switching between different products of the same active substance can have 27 an impact on adherence to medication, because variation in packaging and 28 pill appearance may reduce adherence, especially for chronic diseases.[9, 10] 29 There is a perception among patients and physicians alike that frequent 30 changes between branded and unbranded products (as well as between 31 generics), all containing the same active substance, and especially if patients 32 are older and on multi-drug regimens, may cause patients to become anxious 33 when the appearance of their drugs changes.[11-13] This can lead to an
34 increased risk of patients making mistakes or double medicating, which flows http://bmjopen.bmj.com/ 35 on to increased drug non-adherence.[14, 15] 36 Few studies have investigated clinical differences related to switching among 37 different products of the same active substance in the cardiovascular setting. 38 Until now, most research has focused only on comparing brand-name and 39 generic drugs.[16-20] 40 41 Objectives 3 State specific objectives, including any prespecified hypotheses
42 on September 24, 2021 by guest. Protected copyright. 43 The objectives are stated in the introduction (paragraph 5) of the main 5 44 manuscript: 45 46 The aim of the present study was to investigate the effects of switching to 47 different products of the same off-patent active substance (brand-name or 48 generic) on therapy discontinuation or substitution with another molecule of 49 the same class, in patients with CVD treated with statins and 50 antihypertensives in clinical practice. 51 52 Methods 53 54 Study design 4 Present key elements of study design early in the paper 55 56 The key elements are presented in the method section of the main 57 manuscript: 58 59 60 2 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 27 of 34 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 Data collection 4 5 The data used for the analysis were obtained from the administrative 6 databases of three Local Health Units (LHUs), whose databases included a 7 total population of about two million health-assisted individuals, in the 5,6 8 Italian regions of Lombardy, Lazio and Campania. We analyzed the following 9 archives: Assisted Subjects’ Database, containing the personal data of 10 patients; Medication Prescription database, containing all the information 11 relating to individual prescriptions, such as the International Nonproprietary 12 Names (INN) for pharmaceutical substances, the Anatomical-Therapeutic- 13 Chemical (ATC) code of the prescribed drug, the number of packages, the 14 number of units per package, the dose, the cost per unit and the date of the prescription; Hospital Discharge Database (SDO), containing information on 15 Foreach peer hospital discharge, review in particular the date only of admission and discharge, 16 primary and secondary diagnoses coded according to the International 17 Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM). 18 The patient code in each database allowed electronic linking among all 19 databases. To guarantee patient privacy, this patient code was transcoded 20 into an anonymous univocal numeric code. No identifiers related to patients 21 were provided to the researchers. According to the Italian law for 22 confidentiality of data[22] the study was notified to the Ethic Committees of 23 each LHU. 24 25 Cohort definition 26 27 The study was a retrospective cohort study including all patients aged ≥18 28 years that, between January 1st 2010 and December 31st 2010 (enrollment 29 period), had at least one prescription of simvastatin [ATC code: C10AA01], 30 ramipril [ATC code: C09AA05] or amlodipine [ATC code: C08CA01]. The date 31 of enrollment was defined as the earliest date within the enrollment period 32 in which the patient had the last switch of medication (switchers) or the last 33 prescription in the case of a patient continuing with the same medication
34 (non-switchers). Starting from this date, the individual patient was followed http://bmjopen.bmj.com/ 35 for two years (follow-up period). Switchers were defined as those patients 36 who switched among different products of the same off-patent active 37 substance (i.e., from brand-name to generic, from generic to brand-name, or 38 from generic to another generic). The changes in dose and dosage form were 39 not accounted for during switching. The definitions of the patient cohorts are 40 shown in Figure 1. Data on baseline characteristics, including demographics, 41 cardiovascular risk factors and polymedication, were collected; specifically, polymedication were identified as more than one cardiovascular medication 6,7 42 on September 24, 2021 by guest. Protected copyright. 43 at the date of enrollment. The data on drug prescriptions and 44 hospitalizations that occurred during the 12 months preceding the date of 45 enrollment were analyzed (characterization period). 46 Only patients with at least one prescription of the index drug in the previous 47 12 months were included (to capture patients who could have made a 48 change in therapy) and with at least two prescriptions at follow-up (to 49 include patients with continuity of treatment). Patients treated with fixed 50 combinations of the molecules under consideration (ramipril and diuretics 51 [ATC code: C09BA05], perindopril and amlodipine [ATC code: C09BB04], 52 ramipril and felodipine [ATC code: C09BB05], olmesartan and amlodipine 53 [ATC code: C09DB02], simvastatin and ezetimibe [ATC code: C10BA02]) were 54 excluded. 55 56 Patients transferred to another LHU during the follow-up period were 57 excluded from the analysis. 58 59 60 3 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 28 of 34 BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 Study population 4 5 Cardiovascular risk 6 Patients were classified as high cardiovascular risk if they had cardiovascular 7 treatment or hospitalization for diabetes. Hospitalizations related to diabetes 8 were identified by ICD-9-CM code: 250 (primary discharge reasons); and/or 9 cardiovascular risk factors (previous hospitalization for Ischemic Heart 10 Disease [Acute myocardial infarction (ICD-9-CM: 410) Acute Cardiac Ischemia 11 (ICD-9-CM: 411), Angina Pectoris (ICD-9-CM: 413), Chronic Cardiac Ischemia 12 (ICD-9-CM: 414)]; Cerebral Hemorrhage [ICD-9-CM: 431]; Cerebral Artery 13 Occlusion [ICD-9-CM: 434]; Transient Cerebral Ischemia [ICD-9-CM: 435]; 14 Cerebral Circulatory Disorders [ICD-9-CM: 436]; Atherosclerosis [ICD-9-CM: 15 For440]; peer Other Peripheral review Vascular Disease [ICD-9-CM: only 443]; Chronic Renal 16 Failure [ICD-9-CM: 585]; Coronary Angioplasty [ICD-9-CM procedure: 0066, 17 360]); and/or at least two prescriptions of antidiabetic drugs [ATC code: 18 A10]. All other patients were classified as at moderate cardiovascular risk. 19 20 Drug treatments 21 Patients were also characterized by strategy of treatment at baseline with 22 23 lipid-lowering drugs (ATC code: C10) and antihypertensive drugs (ATC codes: 24 C02, C03, C07, C08, C09). 25 Data analysis at follow-up 26 27 During the follow-up period, the discontinuation or the first substitution of 28 therapy was identified. Discontinuation of therapy was defined as the 29 absence of prescriptions of the same therapeutic class (ATC group) as the 30 index molecule in the last quarter of observation. A substitution of therapy 31 was defined as a change to a different active substance of the same 32 therapeutic class (ATC group) (i.e., switching from simvastatin to a different 7,8 33 statin). A switch among different products of the same active substance was 34 identified by INN. http://bmjopen.bmj.com/ 35 36 Setting 5 Describe the setting, locations, and relevant dates, including periods of 5 37 recruitment, exposure, follow-up, and data collection 38 39 The setting, locations and relevant dates are presented in the method 40 section (paragraphs “Data collection”, “Cohort definition” and “Study 41 population”) of the main manuscript:
42 on September 24, 2021 by guest. Protected copyright. 43 The data used for the analysis were obtained from the administrative 44 databases of three Local Health Units (LHUs), whose databases included a 45 total population of about two million health-assisted individuals, in the 46 Italian regions of Lombardy, Lazio and Campania. 47 48 Participants 6 (a) Give the eligibility criteria, and the sources and methods of selection of NA 49 participants. Describe methods of follow-up 50 51 (b) For matched studies, give matching criteria and number of exposed and 52 unexposed 53 54 The participants are described in the methods section (paragraphs “Cohort 55 definition” and “Study population”) of the main manuscript: 56
57 The study was a retrospective cohort study including all patients aged ≥18 58 years that, between January 1st 2010 and December 31st 2010 (enrollment 59 60 4 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 29 of 34 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 period), had at least one prescription of simvastatin [ATC code: C10AA01], 6 4 ramipril [ATC code: C09AA05] or amlodipine [ATC code: C08CA01]. (…) Only 5 patients with at least one prescription of the index drug in the previous 12 6 months were included (to capture patients who could have made a change in 7 therapy) and with at least two prescriptions at follow-up (to include patients 8 with continuity of treatment). Patients treated with fixed combinations of the 9 molecules under consideration (ramipril and diuretics [ATC code: C09BA05], 7 10 perindopril and amlodipine [ATC code: C09BB04], ramipril and felodipine 11 [ATC code: C09BB05], olmesartan and amlodipine [ATC code: C09DB02], 12 simvastatin and ezetimibe [ATC code: C10BA02]) were excluded.
13 Patients transferred to another LHU during the follow-up period were 14 excluded from the analysis. (…) Patients were classified as high 15 Forcardiovascular peer risk if reviewthey had cardiovascular treatmentonly or hospitalization for 16 diabetes. (…) All other patients were classified as at moderate cardiovascular 17 risk. 18
19 The methods of follow-up are described in the methods section 20 (paragraphs “Cohort definition” and “Study population”) of the main 21 manuscript: 22 23 (…) the individual patient was followed for two years (follow-up period). 24 7 25 Switchers were defined as those patients who switched among different products of the same off-patent active substance (i.e., from brand-name to 26 generic, from generic to brand-name, or from generic to another generic). 27 (…) During the follow-up period, the discontinuation or the first substitution 28 of therapy was identified. Discontinuation of therapy was defined as the 29 absence of prescriptions of the same therapeutic class (ATC group) as the 30 index molecule in the last quarter of observation. A substitution of therapy 31 was defined as a change to a different active substance of the same 32 therapeutic class (ATC group) (i.e., switching from simvastatin to a different 33 statin). A switch among different products of the same active substance was 34 http://bmjopen.bmj.com/ identified by INN. 35 36 37 38 6 39 40 41
42 on September 24, 2021 by guest. Protected copyright. 43 44 8 45 46 47
48 49
50 51
52 53 54 55 56 57 Variables 7 Clearly define all outcomes, exposures, predictors, potential confounders, 58 59 60 5 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 30 of 34 BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 and effect modifiers. Give diagnostic criteria, if applicable 4 5 The outcomes are described in the methods section of the main 6 manuscript: 7 8 8 During the follow-up period, the discontinuation or the first substitution of 9 therapy was identified. Discontinuation of therapy was defined as the 10 absence of prescriptions of the same therapeutic class (ATC group) as the 11 index molecule in the last quarter of observation. A substitution of therapy 12 was defined as a change to a different active substance of the same 13 therapeutic class (ATC group) (i.e., switching from simvastatin to a different 14 statin). A switch among different products of the same active substance was 15 Foridentified peer by INN. review only 16 17 Data sources/ 8* For each variable of interest, give sources of data and details of methods of 18 measurement assessment (measurement). Describe comparability of assessment methods 19 if there is more than one group 20 21 Data sources and details of methods of assessment are given in the 22 methods section first paragraph:
23 24 The data used for the analysis were obtained from the administrative
25 databases of three Local Health Units (LHUs), whose databases included a total population of about two million health-assisted individuals, in the 26 5 27 Italian regions of Lombardy, Lazio and Campania. (...) The study was a retrospective cohort study including all patients aged ≥18 years that, 28 29 between January 1st 2010 and December 31st 2010 (enrollment period), had at least one prescription of simvastatin [ATC code: C10AA01], ramipril [ATC 30 31 code: C09AA05] or amlodipine [ATC code: C08CA01]. (…) the discontinuation 32 or the first substitution of therapy was identified. Discontinuation of therapy 33 was defined as the absence of prescriptions of the same therapeutic class
34 (ATC group) as the index molecule in the last quarter of observation. A http://bmjopen.bmj.com/ 35 substitution of therapy was defined as a change to a different active 36 substance of the same therapeutic class (ATC group) (i.e., switching from 37 simvastatin to a different statin). A switch among different products of the 38 same active substance was identified by INN. 8 39 40 41
42 Bias 9 Describe any efforts to address potential sources of bias NA on September 24, 2021 by guest. Protected copyright. 43 44 This is addressed by the design of the study and the clearly defined data 45 selection specifications 46 47 Study size 10 Explain how the study size was arrived at NA 48 49 How the study size was arrived at is described in the methods section 50 51 Sample size dictated by the source databases 52 53 Quantitative 11 Explain how quantitative variables were handled in the analyses. If 54 variables applicable, describe which groupings were chosen and why 55 56 The handling of the variables is described in the “Statistical analysis” 57 paragraph: 8 58 59 60 6 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 31 of 34 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 Continuous variables were reported as mean ± standard deviation (SD) and 4 compared using the Student's t test; categorical variables were reported as 5 absolute numbers and percentages and compared using the chi-square test. 6 7 Discontinuations of therapy and substitution with another molecule of the 8 same class were analyzed by multivariate analysis using Cox proportional 9 hazards models; covariates considered in the models were: age, male sex, 10 high cardiovascular risk, cardiovascular treatments, change of formulation in 11 the period of characterization. 12 The analysis of Schoenfeld residuals (scaled and unscaled) was conducted to 13 assess the proportionality of risk. 14 15 Forp peer Values <0.05 were review considered statistically significant.only All analyses were 16 performed using STATA 12.0 SE. 17 18 Statistical 12 (a) Describe all statistical methods, including those used to control for 8 19 methods confounding 20 21 (b) Describe any methods used to examine subgroups and interactions 8 22 23 (c) Explain how missing data were addressed NA 24 25 (d) If applicable, explain how loss to follow-up was addressed NA 26 27 (e) Describe any sensitivity analyses 8 28 29 The statistical methods are given in the “Statistical analysis” paragraph: 30 31 Continuous variables were reported as mean ± standard deviation (SD) and 32 compared using the Student's t test; categorical variables were reported as 33 absolute numbers and percentages and compared using the chi-square test.
34 http://bmjopen.bmj.com/ 35 Discontinuations of therapy and substitution with another molecule of the 36 same class were analyzed by multivariate analysis using Cox proportional 37 hazards models; covariates considered in the models were: age, male sex, 38 high cardiovascular risk, cardiovascular treatments, change of formulation in 39 the period of characterization. 40 The analysis of Schoenfeld residuals (scaled and unscaled) was conducted to 41 assess the proportionality of risk.
42 on September 24, 2021 by guest. Protected copyright. 43 p Values <0.05 were considered statistically significant. All analyses were 44 performed using STATA 12.0 SE. 45 46 Results 47 48 Participants 13* (a) Report numbers of individuals at each stage of study—eg numbers 49 potentially eligible, examined for eligibility, confirmed eligible, included in 50 51 the study, completing follow-up, and analysed 52 The numbers of individuals are given in the Results section, paragraphs 53 “Simvastatin”, “Ramipril” and “Amlodipine” 54 9, 13, 55 56 (b) Give reasons for non-participation at each stage NA 57 58 59 60 7 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 32 of 34 BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 (c) Consider use of a flow diagram NA 4 5 Descriptive data 14* (a) Give characteristics of study participants (eg demographic, clinical, social) 6 and information on exposures and potential confounders 7 8 The characteristics of study participants are given in Table 1. 9 9 10 11 (b) Indicate number of participants with missing data for each variable of NA 12 interest 13 14 (c) Summarise follow-up time (eg, average and total amount) 15 For peer review only 16 (…) followed for two years (follow-up period) 6 17 18 Outcome data 15* Report numbers of outcome events or summary measures over time 19 20 Numbers of outcome events were reported in the Results section, 8,9 21 paragraphs “Simvastatin”, “Ramipril” and “Amlodipine” and in Tables 2 10,11 22 and 3 23 24 Main results 16 (a) Give unadjusted estimates and, if applicable, confounder-adjusted 25 estimates and their precision (eg, 95% confidence interval). Make clear 26 which confounders were adjusted for and why they were included 27 28 Variables defined in Statistical analysis section and analyses reported in 29 Table 4 30 8 31 11,12 32 33 (b) Report category boundaries when continuous variables were categorized NA
34 http://bmjopen.bmj.com/ 35 (c) If relevant, consider translating estimates of relative risk into absolute NA 36 risk for a meaningful time period 37 38 Other analyses 17 Report other analyses done—eg analyses of subgroups and interactions, and 39 sensitivity analyses 40 41 Multivariate analysis is described in the “Statistical analysis” paragraph
42 and results are given in Table 4: 8, on September 24, 2021 by guest. Protected copyright. 43 11,12 44 Discontinuations of therapy and substitution with another molecule of the 45 same class were analyzed by multivariate analysis using Cox proportional 46 hazards models; covariates considered in the models were: age, male sex, 47 high cardiovascular risk, cardiovascular treatments, change of formulation in 48 the period of characterization. 49 50 Discussion 51 52 Key results 18 Summarise key results with reference to study objectives 53 54 Key results are summarized as follows in the Discussion: 16,17 55 56 (…) our study indicates that in a “real-world” setting, changes among 57 different product of the same active substance, including switching brand- 58 name to generic, generic to another generic and generic to brand-name, are 59 60 8 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 33 of 34 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 quite common among patients with CVD. Our findings suggest that switching 4 to a different product of the same off-patent active substance, brand-name 5 or generic, may expose patients to a higher risk of therapy discontinuation or 6 substitution than continuing treatment with the same product. 7 8 Limitations 19 Discuss limitations of the study, taking into account sources of potential bias 9 or imprecision. Discuss both direction and magnitude of any potential bias 10 11 Limitations are summarized as follows in the Discussion: 12 16 13 Our analysis has several limitations inherent to any observational study. 14 First, the study was performed using the administrative databases, and the 15 Forreasons peer for switch, non-adherencereview or discontinuation only of treatment in the 16 patients were not retrievable from the dataset. Also, no information on the 17 role of the counseling pharmacist when substituting and dispensing drug 18 packages were available to us. A second limitation is a relatively limited 19 sample size. Although in our study we used the healthcare databases of 20 Lombardy, Lazio and Campania, three Italian Regions localized from north to 21 south of Italy, including data for a total population of about 2 million, and 22 considering that we have focused our analysis among users of simvastatin, 23 ramipril, and amlodipine, larger studies are needed to confirm and to enhance the generalizability of the findings, and in different populations. 24 25 Interpretation 20 Give a cautious overall interpretation of results considering objectives, 26 limitations, multiplicity of analyses, results from similar studies, and other 27 relevant evidence 28 29 The interpretation is given as follows in the Discussion: 30 31 In accordance with previous studies,[23-26] this retrospective analysis in a 14-16 32 “real-world” setting shows that age, gender, cardiovascular risk, and 33 polypharmacy affect discontinuation of therapy. (…) The data of this study 34 show also that for the same active substance, a change of product http://bmjopen.bmj.com/ 35 (regardless of whether it is a brand-name or generic drug) increases the risk 36 of discontinuation of therapy and of substitution with another molecule of 37 the same class. (…) we cannot exclude the possibility that other potential 38 determinants can play a key role in a reduction of patients’ adherence. 39 Observational studies [33, 34] have demonstrated a relationship between 40 age, gender, cardiovascular risk factors, polymedication and a suboptimal 41 adherence to therapy; our findings are in agreement with these previous
42 analyses. In contrast, new evidence suggested that older age is not related to on September 24, 2021 by guest. Protected copyright. 43 poorer medication adherence to cardiovascular medication. A recent 44 systematic search of the bibliographic database MedLine and all Cochrane 45 databases, analyzing the relationship between age and medication 46 adherence in adult patients with chronic heart failure (CHF), showed that 47 that older age alone is not related to poorer medication adherence 48 compared with younger patients with CHF.[35] Our study does not support 49 this concern. 50 However, the underlying reasons for poor adherence are not fully 51 understood, and there may be many reasons behind these behaviours, some 52 of which relate to the perceptions that physicians, pharmacists and patients 53 may have of drugs and therapies. Nevertheless, the clinical consequences 54 that may result from these perceptions are important and should be 55 considered. 56 57 Generalisability 21 Discuss the generalisability (external validity) of the study results 58 59 60 9 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 34 of 34 BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 The issue of generalizability is addressed as follows in the Discussion: 16,17 4 5 Although in our study we used the healthcare databases of Lombardy, Lazio 6 and Campania, three Italian Regions localized from north to south of Italy, 7 including data for a total population of about 2 million, and considering that 8 we have focused our analysis among users of simvastatin, ramipril, and 9 amlodipine, larger studies are needed to confirm and to enhance the 10 generalizability of the findings, and in different populations. 11 12 Despite these limitations, our study indicates that in a “real-world” setting, changes among different product of the same active substance, including 13 switching brand-name to generic, generic to another generic and generic to 14 brand-name, are quite common among patients with CVD. Our findings 15 Forsuggest peer that switching review to a different product ofonly the same off-patent active 16 substance, brand-name or generic, may expose patients to a higher risk of 17 therapy discontinuation or substitution than continuing treatment with the 18 same product. 19 20 21 Other information 22 23 Funding 22 Give the source of funding and the role of the funders for the present study 24 and, if applicable, for the original study on which the present article is based 25 26 Such information is given in the Acknowledgments: 18 27 28 Funding: 29 This research received no specific grant from any funding agency in the 30 public, commercial or not-for-profit sectors. 31 32 *Give information separately for exposed and unexposed groups. 33
34 http://bmjopen.bmj.com/ 35 36 Note: An Explanation and Elaboration article discusses each checklist item and gives methodological background and 37 published examples of transparent reporting. The STROBE checklist is best used in conjunction with this article (freely 38 available on the Web sites of PLoS Medicine at http://www.plosmedicine.org/, Annals of Internal Medicine at 39 http://www.annals.org/, and Epidemiology at http://www.epidem.com/). Information on the STROBE Initiative is 40 41 available at http://www.strobe-statement.org.
42 on September 24, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 10 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
Therapy discontinuation or substitution in patients with cardiovascular disease, switching among different products of the same off-patent active substance: a “real-world” retrospective cohort study
For peer review only Journal: BMJ Open
Manuscript ID bmjopen-2016-012003.R2
Article Type: Research
Date Submitted by the Author: 21-Jul-2016
Complete List of Authors: Degli Esposti, Luca; CliCon S.r.l, Health, Economics and Outcomes Research, Sangiorgi, Diego; CliCon S.r.l. Health, Economics and Outcomes Research Buda, Stefano; CliCon S.r.l. Health, Economics and Outcomes Research Degli Esposti, Ezio; CliCon S.r.l. Health, Economics and Outcomes Research Scaglione, Francesco ; University of Milan, Medical Biotechnology and Translational Medicine
Primary Subject Cardiovascular medicine Heading:
Secondary Subject Heading: Public health http://bmjopen.bmj.com/ Keywords: Simvastatin, Amlodipine, Ramipril, Switching, Adherence to treatment
on September 24, 2021 by guest. Protected copyright.
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 1 of 35 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 5 Therapy discontinuation or substitution in patients with cardiovascular disease, 6 7 switching among different products of the same off-patent active substance: a “real- 8 9 world” retrospective cohort study 10 11 12 13 14 Luca Degli Esposti,1 Diego Sangiorgi,1 Stefano Buda,1 Ezio Degli Esposti,1 Francesco 15 For peer review only 16 2 17 Scaglione 18 19 20 21 22 1CliCon S.r.l. Health, Economics and Outcomes Research, Ravenna, Italy 23 24 2 25 Department of Medical Biotechnology and Translational Medicine, University of Milan, 26 27 Milan, Italy 28 29 30 31 32 Correspondence to 33
34 http://bmjopen.bmj.com/ 35 Luca Degli Esposti, EconD., CliCon S.r.l, Health, Economics and Outcomes Research, Via 36 37 Salara, 36 48100 Ravenna, Italy 38 39 40 Tel +39 544 38393 Fax +39 544 212699. [email protected] 41
42 on September 24, 2021 by guest. Protected copyright. 43 44 45 Keywords 46 47 48 Simvastatin, Amlodipine, Ramipril, Switching, Adherence to treatment 49 50 51 52 53 Word count: 2,452 54 55 56 57 58 59 1 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 2 of 35 BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 ABSTRACT 5 6 Objective: The present study investigated the effects of switching to different products of the 7 8 same off patent active substance (brand name or generic) on therapy discontinuation or 9 10 11 substitution with another molecule of the same class, in patients with cardiovascular disease 12 13 treated with statins and antihypertensives in a “real world” setting. 14 15 For peer review only 16 Design: A retrospective cohort study in a “real world” setting. 17 18 Setting: Analysis of data performed by integrating administrative databases that included 19 20 21 approximately two million individuals who are assisted by the National Health System from 22 23 three Local Health Units located in three different regions of Italy. 24 25 Participants: All patients aged ≥18 years with at least one prescription of simvastatin, 26 27 st st 28 ramipril or amlodipine in the period January 1 to December 31 2010 were included and 29 30 followed up for two years. 31 32 33 Main outcome measures: Prescription refills occurring during follow up were evaluated.
34 http://bmjopen.bmj.com/ 35 Frequency of discontinuation of therapy or substitution with another molecule of the same 36 37 class (for example, from simvastatin to a different statin) during follow up was identified. 38 39 40 Results: During follow up, therapy discontinuation or substitution was found to be more 41
42 frequent in patients switching to a different product of the same active substance compared on September 24, 2021 by guest. Protected copyright. 43 44 with non switching patients (11.5% vs 10.8% and 22.2% vs 20.8% [p=0.002], respectively, 45 46 in the simvastatin group; 4.0% vs 3.5% and 24.6% vs 22.7% [p<0.001], respectively, in the 47 48 49 amlodipine group). In the ramipril group, 8% of patients undertook a therapy substitution to 50 51 another molecule; no trend towards a lower percentage of substitution was observed in the 52 53 non switching group, while 18% of patients discontinued treatment, with a significant 54 55 difference in favour of patients not switching. These findings were partially confirmed by 56 57 58 multivariate analysis. 59 2 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 3 of 35 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 Conclusions: Switches among products of the same active substance are quite common in 5 6 patients with cardiovascular disease. Our study suggests that switching may expose patients 7 8 to a higher risk of therapy discontinuation or substitution. 9 10 11 12 13 14 STRENGTHS AND LIMITATIONS OF THIS STUDY 15 For peer review only 16 17 • This study, in a “real world” setting, is one of only a few studies to investigate 18 19 clinical differences related to switching among different products of the same active 20 21 substance in the cardiovascular setting. Until now, most research has focused only on 22 23 comparing brand name and generic drugs. 24 25 26 • Sample size was relatively limited, and although we used three healthcare databases 27 28 comprising a total of approximately two million individuals who are assisted by the 29 30 National Health System in three regions of Italy, larger studies are needed to confirm 31 32 33 and to enhance the generalisability of the findings, and in different populations.
34 http://bmjopen.bmj.com/ 35 • In common with other retrospective, observational studies, reasons for switch, non 36 37 38 adherence or discontinuation of treatment were not retrievable from the data set. 39 40 41
42 on September 24, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 3 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 4 of 35 BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 INTRODUCTION 5 6 7 Cardiovascular diseases (CVD) are the leading cause of death worldwide, accounting for 8 9 approximately one third of all deaths.[1] Combination therapy with antihypertensive drugs 10 11 12 and serum cholesterol lowering drugs is effective in prevention, and it is estimated that a high 13 14 level of adherence to treatment will reduce the risk of CVD by approximately 80%.[2] A 15 For peer review only 16 number of studies have demonstrated that patients often discontinue long term treatment or 17 18 take less than prescribed, and that such non adherence reduces the potential preventive 19 20 benefits.[3] 21 22 23 Many reasons contribute to patient non adherence to medical therapy, such as ageing, 24 25 comorbidities, poor relationship between patient and physician, poor memory, and patients’ 26 27 28 low perception of disease severity.[4] In addition, the need to take several drugs 29 30 concomitantly, or other medication related factors, may make remembering when to take 31 32 each drug more difficult and increase the risk of possible side effects caused by adverse drug 33 34 drug interactions. Although medication side effects are probably not the main cause of poor http://bmjopen.bmj.com/ 35 36 adherence, as there seems to be little direct relationship between adherence and drug class,[3] 37 38 39 they are also associated with treatment discontinuation, especially in the early treatment of 40 41 hypertension.[5 8] Kronish et al showed that in the clinical setting adherence to diuretics and
42 on September 24, 2021 by guest. Protected copyright. 43 β blockers is lowest and the highest adherence is to angiotensin II receptor blockers and 44 45 angiotensin converting enzyme inhibitors.[5] Similarly, a retrospective study based on a 46 47 48 cohort of 207,473 patients in Ontario found that treatment with angiotensin converting 49 50 enzyme inhibitors showed the best therapy persistence and compliance, and β blockers 51 52 showed the worst compliance (all p<0.001).[8] Furthermore, some studies have demonstrated 53 54 that switching between different products of the same active substance can have an impact on 55 56 57 58 59 4 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 5 of 35 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 adherence to medication, because variation in packaging and pill appearance may reduce 5 6 adherence, especially for chronic diseases.[9, 10] 7 8 9 There is a perception among patients and physicians alike that frequent changes between 10 11 branded and unbranded products (as well as between generics), all containing the same active 12 13 substance, and especially if patients are older and on multi drug regimens, may cause patients 14 15 For peer review only 16 to become anxious when the appearance of their drugs changes.[11 13] This can lead to an 17 18 increased risk of patients making mistakes or double medicating, which flows on to increased 19 20 drug non adherence.[14, 15] 21 22 23 Few studies have investigated clinical differences related to switching among different 24 25 products of the same active substance in the cardiovascular setting. Until now, most research 26 27 has focused only on comparing brand name and generic drugs.[16 20] 28 29 30 The aim of the present study was to investigate the effects of switching to different 31 32 products of the same off patent active substance (brand name or generic) on therapy 33 34 discontinuation or substitution with another molecule of the same class, in patients with CVD http://bmjopen.bmj.com/ 35 36 37 treated with statins and antihypertensives in a “real world” setting. 38 39 A version of this article has previously been published as a journal supplement in the 40 41 Italian language.[21]
42 on September 24, 2021 by guest. Protected copyright. 43 44 METHODS 45 46 47 Data collection 48 49 50 The data used for the analysis were obtained from the administrative databases of three Local 51 52 Health Units (LHUs), whose databases included a total population of about two million 53 54 individuals who are assisted by the National Health System, in the Italian regions of 55 56 57 Lombardy, Lazio and Campania. We analysed the following archives: Assisted Subjects’ 58 59 5 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 6 of 35 BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 Database, containing the personal data of patients; Medication Prescription database, 5 6 containing all the information relating to individual prescriptions dispensed by the pharmacy, 7 8 such as the International Nonproprietary Names (INN) for pharmaceutical substances, the 9 10 11 Anatomical Therapeutic Chemical (ATC) code of the prescribed drug, the number of 12 13 packages, the number of units per package, the dose, the brand name drug, the cost per unit 14 15 and the date Forof the prescription; peer Hospital review Discharge Database (SDO),only containing information 16 17 on each hospital discharge, in particular the date of admission and discharge, primary and 18 19 20 secondary diagnoses coded according to the International Classification of Diseases, Ninth 21 22 Revision, Clinical Modification (ICD-9-CM). The patient code in each database allowed 23 24 electronic linking among all databases. To guarantee patient privacy, this patient code was 25 26 transcoded into an anonymous univocal numeric code. No identifiers related to patients were 27 28 provided to the researchers. According to the Italian law for confidentiality of data [22] the 29 30 31 study was notified to the Ethic Committees of each LHU. 32 33 Cohort definition 34 http://bmjopen.bmj.com/ 35 36 The study was a retrospective cohort study including all patients aged ≥18 years that, 37 38 between January 1st 2010 and December 31st 2010 (enrollment period), had at least one 39 40 41 prescription of simvastatin [ATC code: C10AA01], ramipril [ATC code: C09AA05] or
42 on September 24, 2021 by guest. Protected copyright. 43 amlodipine [ATC code: C08CA01] as a brand name or generic prescription. The date of 44 45 enrollment was defined as the earliest date within the enrollment period in which the patient 46 47 had the last switch of medication or the last prescription in the case of a patient continuing 48 49 with the same medication. Starting from this date, the individual patient was followed for two 50 51 52 years (follow up period). The patient cohorts were defined in the following way: non 53 54 switchers were defined as those patients who did not change medication, regardless of 55 56 whether it was brand name or generic; switchers were defined as those patients who switched 57 58 59 6 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 7 of 35 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 among different products of the same off patent active substance (i.e., from brand name to 5 6 generic, from generic to brand name, or from generic to another generic). The changes in 7 8 dose and dosage form were not accounted for during switching. Data on baseline 9 10 11 characteristics, including demographics, cardiovascular risk factors and receipt of more than 12 13 one cardiovascular medication at the date of enrolment were collected. The data on drug 14 15 prescriptionsFor and hospitalisations peer that occurredreview during the 12 onlymonths preceding the date of 16 17 enrollment were analysed (characterisation period). The medication adherence in the year 18 19 20 before the index date was also analysed. Adherence to therapy was determined by calculating 21 22 the proportion of days covered according to the method used by Catalan and LeLorier.[23] 23 24 Only patients with at least one prescription of the index drug in the previous 12 months 25 26 27 were included (to capture patients who could have made a change in therapy) and with at 28 29 least two prescriptions at follow up (to include patients with continuity of treatment). Patients 30 31 treated with fixed combinations of the molecules under consideration (ramipril and diuretics 32 33 [ATC code: C09BA05], perindopril and amlodipine [ATC code: C09BB04], ramipril and 34 http://bmjopen.bmj.com/ 35 36 felodipine [ATC code: C09BB05], olmesartan and amlodipine [ATC code: C09DB02], 37 38 simvastatin and ezetimibe [ATC code: C10BA02]) were excluded. 39 40 41 Patients transferred to another LHU during the follow up period were excluded from the
42 on September 24, 2021 by guest. Protected copyright. 43 analysis. 44 45 Study population 46 47 48 Cardiovascular risk 49 50 51 Patients were classified as high cardiovascular risk if they had cardiovascular treatment or 52 53 hospitalisation for diabetes. For each patient, hospitalisations related to diabetes were 54 55 identified by ICD 9 CM code: 250 (primary discharge reasons); and/or cardiovascular risk 56 57 58 factors (previous hospitalisation for Ischemic Heart Disease [Acute myocardial infarction 59 7 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 8 of 35 BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 (ICD 9 CM: 410) Acute cardiac ischemia (ICD 9 CM: 411), Old myocardial infarction (ICD 5 6 9 CM: 412), Angina pectoris (ICD 9 CM: 413), Chronic cardiac ischemia (ICD 9 CM: 414)]; 7 8 Heart failure [ICD 9 CM: 428]; Cerebral Haemorrhage [ICD 9 CM: 431]; Cerebral Artery 9 10 11 Occlusion [ICD 9 CM: 434]; Transient Cerebral Ischaemia [ICD 9 CM: 435]; Cerebral 12 13 Circulatory Disorders [ICD 9 CM: 436]; Atherosclerosis [ICD 9 CM: 440]; Other Peripheral 14 15 Vascular DiseaseFor [ICD 9 CM: peer 443]; Chronicreview Renal Failure only [ICD 9 CM: 585]; Coronary 16 17 Angioplasty [ICD 9 CM procedure: 0066, 360]); and/or the presence of at least two 18 19 20 prescriptions of antidiabetic drugs [ATC code: A10]. All other patients were classified as at 21 22 moderate cardiovascular risk. 23 24 Drug treatments 25 26 27 Patients were also characterised by strategy of treatment at baseline with lipid lowering drugs 28 29 (ATC code: C10) and antihypertensive drugs (ATC codes: C02, C03, C07, C08, C09). 30 31 32 Data analysis at follow up 33
34 http://bmjopen.bmj.com/ 35 During the follow up period, the discontinuation or the first substitution of therapy was 36 37 identified. Discontinuation of therapy was defined as the absence of prescriptions of the same 38 39 therapeutic class (ATC group) as the index molecule in the last quarter of observation. A 40 41 substitution of therapy was defined as a change to a different active substance of the same
42 on September 24, 2021 by guest. Protected copyright. 43 44 therapeutic class (ATC group) (i.e., switching from simvastatin to a different statin). A switch 45 46 among different products of the same active substance was identified by INN. 47 48 49 Statistical analysis 50 51 Continuous variables were reported as mean ± standard deviation (SD) and compared using 52 53 54 the Student's t test; categorical variables were reported as absolute numbers and percentages 55 56 and compared using the chi square test. 57 58 59 8 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 9 of 35 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 Discontinuations of therapy and substitution with another molecule of the same class 5 6 were analysed by multivariate analysis using Cox proportional hazards models; covariates 7 8 considered in the models were: age, male sex, high cardiovascular risk, cardiovascular 9 10 11 treatments, change of formulation in the period of characterisation. 12 13 The analysis of Schoenfeld residuals (scaled and unscaled) was conducted to assess the 14 15 For peer review only 16 proportionality of risk. 17 18 p Values <0.05 were considered statistically significant. All analyses were performed 19 20 21 using STATA 12.0 SE. 22 23 RESULTS 24 25 26 Simvastatin 27 28 29 A total of 38,183 patients treated with simvastatin, 17,642 male (46%), mean age 68.3±10.7 30 31 years, were included in the analysis. A total of 9,392 (25%) patients were classified as at high 32 33 cardiovascular risk, while 30,467 (80%) received concomitant cardiovascular treatments 34 http://bmjopen.bmj.com/ 35 36 (Table 1). 37 38 Table 1 Characteristics of the enrolled patients 39 40 41 Simvastatin Ramipril Amlodipine
42 on September 24, 2021 by guest. Protected copyright. 43 Non- p Non- p Non- p 44 Category switchers Switchers Value switchers Switchers Value switchers Switchers Value 45 46 47 Total 23,180 (61) 15,003 (39) 22,799 (71) 9,312 (29) 26,823 (72) 10,644 (28) 48 49 Patient Characteristics 50 51 Age, y 68.5 ± 10.7 68.0 ± 10.6 <0.001 67.0 ± 12.7 66.8 ± 13.0 0.203 68.3 ± 11.7 68.1 ± 11.8 0.137 52 53 54 Male sex 10,469 (45.2) 7,173 (47.8) <0.001 12,876 5,617 (60.3) <0.001 14,126 6,213 (58.4) <0.001 55 (56.5) (52.7) 56 57 58 59 9 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 10 of 35 BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 High CVR 5,494 (23.7) 3,898 (26.0) <0.001 4,873 (21.4) 2,025 (21.7) 0.470 4,983 (18.6) 2,143 (20.1) <0.001 5 6 7 Additional 18,539 (80.0) 11,928 0.265 17,868 7,393 (79.4) 0.044 23,793 9,588 (90.1) <0.001 8 treatments (79.5) (78.4) (88.7) 9 10 Values are N (%) or mean ± standard deviation (SD). 11 12 13 CVR, cardiovascular risk. 14 15 For peer review only 16 Switches among different products occurred in 39% of patients treated with simvastatin. 17 18 Switcher patients were mainly men with high cardiovascular risk; this cohort was slightly 19 20 younger than non switchers, the difference was statistically significant. With regard to 21 22 switchers, a little over half carried out one switch only during the characterisation period, 23 24 25 with 8% having four switches or more (Table 2). Among patients enrolled, the non switching 26 27 and switching group showed a similarly percentage of adherence during the characterisation 28 29 period [34.2% vs 33.5% (p=0.133), respectively]. 30 31 32 33
34 http://bmjopen.bmj.com/ 35 Table 2 Annual frequency of switches 36 37 38 Simvastatin Ramipril Amlodipine 39 40 41 Switches (N) Patients (%) Switches (N) Patients (%) Switches (N) Patients (%)
42 on September 24, 2021 by guest. Protected copyright. 43 1 7,842 (52) 1 5,112 (55) 1 5,710 (54) 44 45 2 4,062 (27) 2 2,501 (27) 2 2,871 (27) 46 47 48 3 1,917 (13) 3 1,116 (12) 3 1,300 (12) 49 50 4 805 (5) 4 413 (4) 4 563 (5) 51 52 ≥5 377 (3) ≥5 170 (2) ≥5 200 (2) 53 54 55 56 57 58 59 10 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 11 of 35 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 In the follow up period, 4,232 (11%) patients undertook a therapy substitution with another 5 6 molecule; a significantly lower percentage of substitution was observed in the group that did 7 8 not switch to a different product of the same active substance (Table 3). 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33
34 http://bmjopen.bmj.com/ 35 36 37 38 39 40 41
42 on September 24, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 11 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 12 of 35 BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 Table 3 Therapy discontinuation and substitution in switcher and non switcher patients 5 6 7 Simvastatin Ramipril Amlodipine 8 9 Switch Switch Switch 10 11 No Yes p No Yes p No Yes p 12 13 Value Value Value 14 15 Total, N (%) For23,180 (61) 15,003peer (39) review 22,799 (71) 9,312 (29)only 26,823 (72) 10,644 (28) 16 17 18 Therapy substitution or discontinuation 19 20 Substitution, N (%) 2,506 1,726 0.037 1,785 (7.8) 711 (7.6) 0.571 943 (3.5) 426 (4.0) 0.026 21 (10.8) (11.5) 22 23 24 Discontinuation, N (%) 4,826 3,327 0.002 3,885 (17.0) 1,792 (19.2) <0.001 6,089 (22.7) 2,618 (24.6) <0.001 25 (20.8) (22.2) 26 27 28 29 30 In the same period 8,153 (21%) patients discontinued treatment; a significantly lower 31 32 percentage of discontinuation was observed for non switching patients (Table 3). These 33
34 http://bmjopen.bmj.com/ 35 findings were partially confirmed by multivariate analysis (Table 4): the group that switched 36 37 to a different product of the same active substance showed a higher probability of 38 39 discontinuation (hazard ratio [HR]=1.087, 95% CI 1.040 1.136, p<0.001) and a higher, but 40 41 not significant, probability of substitution of therapy (HR=1.059, 95% CI 0.996 1.126,
42 on September 24, 2021 by guest. Protected copyright. 43 44 p=0.068). 45 46 Table 4 Multivariate analysis of predictors of risk of therapy discontinuation and substitution 47 48 49 Variable Simvastatin 50 51 52 Substitution Discontinuation 53 54 HR 95% CI p Value HR 95% CI p Value 55 56 57 Age 0.993 0.990 0.995 <0.001 1.001 0.999 1.004 0.194 58 59 12 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 13 of 35 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 Male sex 1.116 1.050 1.186 <0.001 0.921 0.882 0.963 <0.001 5 6 7 High CVR 1.107 1.034 1.185 0.004 0.962 0.914 1.013 0.140 8 9 Additional CV treatments 1.064 0.984 1.152 0.121 0.821 0.778 0.867 <0.001 10 11 Switch to different product of 12 13 the same substance 1.059 0.996 1.126 0.068 1.087 1.040 1.136 <0.001 14 15 For peer reviewRamipril only 16 17 18 Substitution Discontinuation 19 20 HR 95% CI p Value HR 95% CI p Value 21 22 Age 1.003 1.000 1.006 0.072 1.005 1.002 1.007 <0.001 23 24 25 Male sex 0.938 0.865 1.017 0.119 0.950 0.900 1.002 0.058 26 27 High CVR 1.158 1.057 1.269 0.002 0.866 0.811 0.926 <0.001 28 29 Additional CV treatments 1.521 1.360 1.700 <0.001 0.806 0.757 0.857 <0.001 30 31 32 Switch to different product of 33 the same substance 0.973 0.892 1.062 0.540 1.163 1.100 1.230 <0.001
34 http://bmjopen.bmj.com/ 35 36 Amlodipine 37 38 Substitution Discontinuation 39 40 HR 95% CI p Value HR 95% CI p Value 41
42 on September 24, 2021 by guest. Protected copyright. 43 Age 1.019 1.013 1.024 <0.001 1.006 1.004 1.008 <0.001 44 45 Male sex 0.807 0.719 0.907 <0.001 0.870 0.834 0.909 <0.001 46 47 High CVR 1.247 1.091 1.424 0.001 0.879 0.831 0.929 <0.001 48 49 50 Additional CV treatments 1.935 1.514 2.474 <0.001 0.959 0.897 1.025 0.215 51 52 Switch to different product of 53 54 the same substance 1.179 1.043 1.333 0.008 1.124 1.074 1.177 <0.001 55 56 CI, confidence interval; CV, cardiovascular; CVR, cardiovascular risk; HR, hazard ratio. 57 58 59 13 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 14 of 35 BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 Ramipril 5 6 7 A total of 32,111 patients treated with ramipril, 18,493 male (58%), mean age 66.9±12.8 8 9 years, were included in the analysis. Of these, 6,898 (21%) patients were classified as at high 10 11 cardiovascular risk, while 25,261 (79%) were receiving additional cardiovascular treatments 12 13 (Table 1). Switches among different products occurred in 29% of patients treated with 14 15 For peer review only 16 ramipril. Switcher patients were mainly men and the difference was statistically significant. 17 18 The switching group showed a higher percentage of adherence than non switcher patients 19 20 during the characterisation period [48.9% vs 46.6% (p=0.001), respectively]. 21 22 23 With regard to switchers, again, a little over half (55%) carried out one switch only, during 24 25 the characterisation period, and few (6%) had four switches or more (Table 2). In the follow 26 27 up period, 2,496 (8%) patients undertook a therapy substitution to another molecule; no trend 28 29 towards a lower percentage of substitution was observed in the group that did not switch to a 30 31 32 different product of the same active substance (Table 3). In the same period, 5,677 (18%) 33
34 patients discontinued treatment, with a significant difference in favour of patients not http://bmjopen.bmj.com/ 35 36 switching to a different product of the same active substance (Table 3). These findings were 37 38 confirmed by multivariate analysis (Table 4): there was essentially no difference between 39 40 41 groups in terms of probability of substitution (HR=0.973, 95% CI 0.892 1.062, p=0.540),
42 on September 24, 2021 by guest. Protected copyright. 43 while the non switching group showed a significantly lower probability of discontinuation of 44 45 therapy (HR=1.163, 95% CI 1.100 1.230, p<0.001). 46 47 48 Amlodipine 49 50 A total of 37,467 patients treated with amlodipine, 20,339 male (54%), mean age 68.2±11.7 51 52 53 years, were included in the analysis. Of these, 7,126 (19%) patients were classified as at high 54 55 cardiovascular risk, while 33,381 (89%) were receiving additional cardiovascular treatments 56 57 (Table 1). Switches among different products occurred in 28% of patients treated with 58 59 14 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 15 of 35 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 amlodipine. Switcher patients were mainly men with high cardiovascular risk; the difference 5 6 was statistically significant. Among patients enrolled, switcher patients showed a lower 7 8 percentage of adherence during the characterisation period than non switcher patients [42.2% 9 10 11 vs 43.8% (p=0.007), respectively]. With regard to switchers, just over half (54%) carried out 12 13 one switch only during the characterisation period, and 7% had four switches or more (Table 14 15 2). In the follow upFor period, peer 1,369 (4%) reviewpatients undertook a therapyonly substitution to another 16 17 molecule; a significantly lower percentage of substitution was observed in the group that did 18 19 20 not switch to a different product of the same active substance (Table 3). In the same period, 21 22 8,707 (23%) patients discontinued treatment; a significantly lower probability of 23 24 discontinuation was observed for patients not switching to another product of the same active 25 26 substance (Table 3). These findings were confirmed by multivariate analysis (Table 4): the 27 28 switcher group showed a higher probability of discontinuation (HR=1.124, 95% CI 1.074 29 30 31 1.177, p<0.001) and substitution of therapy (HR=1.179, 95% CI 1.043 1.333, p=0.008). 32 33
34 DISCUSSION http://bmjopen.bmj.com/ 35 36 In accordance with previous studies,[24 28] this retrospective analysis in a “real world” 37 38 39 setting shows that age, gender, cardiovascular risk, and more than one cardiovascular 40 41 medication at the date of enrollment could to play a role in the discontinuation of therapy. A
42 on September 24, 2021 by guest. Protected copyright. 43 number of factors may interact to affect adherence to therapies for chronic conditions. These 44 45 have been categorised by the World Health Organization as social and economic related 46 47 48 factors, health system/health care team related factors, condition related factors, and patient 49 50 related factors.[15] As poor adherence has a significant negative impact on health outcomes 51 52 and healthcare costs, and imposes a substantial burden on patients and health systems, 53 54 “increasing the effectiveness of adherence interventions may have a far greater impact on the 55 56 health of the population than any improvement in specific medical treatments”.[29] 57 58 59 15 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 16 of 35 BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 The data of this study show also that for the same active substance, a change of product 5 6 (regardless of whether it is a brand name or generic drug) increases the risk of 7 8 discontinuation of therapy and of substitution with another molecule of the same class. Our 9 10 11 findings were confirmed by multivariate analysis, where the switcher group showed a higher 12 13 probability of discontinuation and probability of substitution for amlodipine and simvastatin 14 15 users. Instead,For among ramipril peer users there review was essentially no difference only between groups in 16 17 terms of probability of substitution while a higher probability of discontinuation was 18 19 20 confirmed by multivariate analysis. 21 22 Our findings are comparable with those reported by Ghate et al.,[30] who found that 23 24 switching among warfarin formulations, including substituting a generic for another generic, 25 26 27 might expose patients with atrial fibrillation to a higher risk of thrombotic and bleeding 28 29 events than remaining on the same formulation. 30 31 32 There is a lot of published evidence about switching from branded to generic medicines, 33
34 specifically regarding the role of prescribers and pharmacists in the opportunity for generic http://bmjopen.bmj.com/ 35 36 drug use and generic substitution, as well as concerning the acceptance by patients of generic 37 38 substitution by health providers.[31, 32] According to Italian law (Patent Law and the Health 39 40 41 Law Regulations in Italy. Decree 95/2012), all pharmacists in Italy are required to offer
42 on September 24, 2021 by guest. Protected copyright. 43 patients the opportunity to substitute a prescribed non generic, interchangeable medicinal 44 45 product with a less expensive generic alternative, unless the prescriber states specifically that 46 47 the prescription is non substitutable. At the same time, the patient can decline the substitution 48 49 of a medicinal product. 50 51 52 However, a previous study exploring the effect of generic substitution showed that 53 54 physician induced switching from brand name to generic ramipril does not negatively affect 55 56 57 the refill compliance of patients.[16] 58 59 16 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 17 of 35 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 In contrast, at present only few studies have estimated the frequency and effects of 5 6 substitution between different products of the same active substance in a clinical practice 7 8 setting. Previous analyses suggest that patients switching statin therapy showed significantly 9 10 11 poorer compliance and higher risk of death or major cardiovascular events when compared 12 13 with controls who did not switch.[33 35] 14 15 For peer review only 16 Moreover, as observed in other studies, this study indicates that age, sex and the presence 17 18 of cardiovascular risk were associated significantly with the presence of switching to a 19 20 different product of the same active substance.[36] 21 22 23 In addition, our results are also comparable with others that focused on different chronic 24 25 therapies, such as a recent study by Kesselheim et al. showing that changes in pill colours and 26 27 shapes increased the risk of non adherence among epileptic patients.[37] The possibility that 28 29 variation in packaging and pill appearance, as well as in the shape and colour of either box or 30 31 32 tablet, may affect adherence is a reason for concern. 33 34 However, we cannot exclude the possibility that other potential determinants can play a http://bmjopen.bmj.com/ 35 36 37 key role in a reduction of patients’ adherence. Observational studies [38 40] have 38 39 demonstrated a relationship between age, gender, cardiovascular risk factors, more than one 40 41 cardiovascular medication and a suboptimal adherence to therapy; our findings are in
42 on September 24, 2021 by guest. Protected copyright. 43 agreement with these previous analyses. The majority of the published studies showed that 44 45 age was related to adherence, although a few researchers found age not to be a factor causing 46 47 48 non adherence. New evidence suggests that older age is not related to poorer medication 49 50 adherence to cardiovascular medication. A recent systematic search of the bibliographic 51 52 database MedLine and all Cochrane databases, analysing the relationship between age and 53 54 medication adherence in adult patients with chronic heart failure (CHF), showed that older 55 56 57 age alone is not related to poorer medication adherence compared with younger patients with 58 59 17 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 18 of 35 BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 CHF.[41] Our study does not support this concern. Several studies also attempted to 5 6 hypothesise plausible reasons for poorer compliance among elderly patients. Elderly patients 7 8 may have problems with vision, hearing and memory. In addition, they may have more 9 10 11 difficulties in following therapy instructions due to cognitive impairment or other physical 12 13 difficulties, such as having problems in swallowing tablets, opening drug containers, handling 14 15 small tablets,For distinguishing peer colours or identifyingreview markings on only drugs.[42] 16 17 18 However, the underlying reasons for poor adherence are not fully understood, and there may 19 20 be many reasons behind these behaviours, some of which relate to the perceptions that 21 22 physicians, pharmacists and patients may have of drugs and therapies. Nevertheless, the 23 24 clinical consequences that may result from these perceptions are important and should be 25 26 27 considered. Establishing better physician patient communication, improving patient 28 29 education, and maintaining regular follow up and review of patients’ progress may be as 30 31 important as other factors in encouraging adherence and lead to improved health outcomes 32 33 and enhanced patient safety.[15] This includes addressing patients’ perceptions about the 34 http://bmjopen.bmj.com/ 35 36 medications they are prescribed and understanding that they may find routine changes in the 37 38 name and appearance of long term medications challenging. 39 40 41 Our analysis has several limitations inherent to any observational study. First, the study was
42 on September 24, 2021 by guest. Protected copyright. 43 performed using the administrative databases, and the reasons for switch, non adherence or 44 45 discontinuation of treatment in the patients were not retrievable from the dataset. Also, no 46 47 information on the role of the prescribers regarding switching within the same class or the 48 49 role of the counselling pharmacist when substituting and dispensing drug packages were 50 51 52 available to us. A second limitation is a relatively limited sample size. Although in our study 53 54 we used the healthcare databases of Lombardy, Lazio and Campania, three Italian Regions 55 56 localised from north to south of Italy, including data for a total population of about 2 million, 57 58 59 18 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 19 of 35 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 and considering that we have focused our analysis among users of simvastatin, ramipril, and 5 6 amlodipine, larger studies are needed to confirm and to enhance the generalisability of the 7 8 findings, and in different populations. Thirdly, our study did not include an outcome analysis, 9 10 11 and the evaluation of the clinical consequences of switching was beyond the scope of this 12 13 work. 14 15 For peer review only 16 Despite these limitations, our study indicates that in a “real world” setting, changes among 17 18 different product of the same active substance, including switching brand name to generic, 19 20 generic to another generic and generic to brand name, are quite common among patients with 21 22 CVD. Our findings suggest that switching to a different product of the same off patent active 23 24 substance, brand name or generic, may expose patients to a higher risk of therapy 25 26 27 discontinuation or substitution than continuing treatment with the same product. 28 29 30 31 32 33
34 http://bmjopen.bmj.com/ 35 36 37 38 39 40 41
42 on September 24, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 19 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 20 of 35 BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 Acknowledgements: We thank Ray Hill and Gayle Robins, independent medical writers, 5 6 who provided English language editing and journal styling prior to submission. 7 8 9 Contributors: LDE, FS, DS conceived and designed the study; LDE, DS, SB, EDE analysed 10 11 the data; and LDE, FS, DS wrote the paper. All authors revised and approved the final 12 13 version. 14 15 For peer review only 16 Funding: This research received no specific grant from any funding agency in the public, 17 18 commercial or not for profit sectors. 19 20 Competing interests: All authors have completed the Unified Competing Interest form at 21 22 http://www.icmje.org/coi_disclosure.pdf (available on request from the corresponding author) 23 24 and declare no support from any organisation for the submitted work. 25 26 27 FS has served as consultant and had speaking engagements to pharma companies marketing 28 29 cardiovascular drugs and has been compensated for travel and time spent on research and 30 31 lectures. 32 33 Data sharing statement: No additional data are available. 34 http://bmjopen.bmj.com/ 35 36 Ethics approval: For this type of study, formal consent is not required. However, to 37 38 guarantee patient privacy, no personal identifiers were provided to the researchers. According 39 40 to the Italian law for confidentiality of data, the study was notified to the Ethic Committees of 41
42 each Local Health Unit. on September 24, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 20 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 21 of 35 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 References 5 6 7 1. World Health Organization. The top 10 causes of death. 2014. Available at: 8 9 http://www.who.int/mediacentre/factsheets/fs310/en/index.html (accessed September 30, 10 11 2014) 12 13 2. Wald NJ, Law MR. A strategy to reduce cardiovascular disease by more than 80%. BMJ 14 15 2003;326:1419 23.For peer review only 16 17 18 3. Naderi SH, Bestwick JP, Wald DS. Adherence to drugs that prevent cardiovascular 19 20 disease: meta analysis on 376,162 patients. Am J Med 2012;125:882 7.e1 21 22 4. Burke LE, Dunbar Jacobs JM, Hill MN. Compliance with cardiovascular disease 23 24 prevention strategies: a review of the research. Ann Behav Med 2012;19:239 63. 25 26 27 5. Kronish IM, Woodward M, Sergie Z, et al. Meta analysis: impact of drug class on 28 29 adherence to antihypertensives. Circulation 2011;123:1611 21. 30 31 32 6. Burke TA, Sturkenboom MC, Lu SE, et al. Discontinuation of antihypertensive drugs 33
34 among newly diagnosed hypertensive patients in UK general practice. J Hypertens http://bmjopen.bmj.com/ 35 36 2006;24:1193 200. 37 38 39 7. Erkens JA, Panneman MM, Klungel OH, et al. Differences in antihypertensive drug 40 41 persistence associated with drug class and gender: a PHARMO study.
42 on September 24, 2021 by guest. Protected copyright. 43 Pharmacoepidemiol Drug Saf 2005;14:795 803. 44 45 46 8. Friedman O, McAlister FA, Yun L, et al.; Canadian Hypertension Education Program 47 48 Outcomes Research Taskforce. Antihypertensive drug persistence and compliance among 49 50 newly treated elderly hypertensives in Ontario. Am J Med 2010;123:173 81. 51 52 53 9. Gabbay U, Yosef N, Feder Krengel N, et al. Therapeutic equivalent substitute that is new 54 55 or unfamiliar to the chronic patient may result in medication error. Int J Health Care 56 57 Qual Assur 2012;25:509 18. 58 59 21 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 22 of 35 BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 10. Greene JA. The substance of the brand. Lancet 2011;378:120 1. 5 6 11. Posner J, Griffin JP. Generic substitution. Br J Clin Pharmacol 2011;72:731 2. 7 8 12. Greene JA, Kesselheim AS. Why do the same drugs look different? Pills, trade dress, and 9 10 11 public health. N Engl J Med 2011;365:83 9. 12 13 13. Hakonsen H, Eilertsen M, Borge H, et al. Generic substitution: additional challenge for 14 15 adherenceFor in hypertensive peer patients? Currreview Med Res Opin 2009;25:2515 21. only 16 17 14. Atar D, Carmena R, Clemmensen P, et al. Clinical review: impact of statin substitution 18 19 20 policies on patient outcomes. Ann Med 2009;41:242 56. 21 22 15. World Health Organization. Adherence to long term therapies: evidence for action. 23 24 2003.WHO, Geneva, Switzerland. Available from 25 26 http://www.emro.who.int/ncd/Publications/adherence_report.pdf (accessed September 27 28 30, 2014). 29 30 31 16. Ude M, Schuessel K, Quinzler R, et al. Generic switch after ramipril patent expiry is not 32 33 associated with decreased pharmacy refill compliance: a retrospective study using the
34 http://bmjopen.bmj.com/ 35 DAPI database. J Hypertens 2011;29:1837 45. 36 37 17. Van Wijk BL, Klungel OH, Heerdink ER, et al. Generic substitution of antihypertensive 38 39 40 drugs: does it affect adherence? Ann Pharmacother 2006;40:15 20. 41
42 18. Chapman RH, Benner JS, Girase P, et al. Generic and therapeutic statin switches and on September 24, 2021 by guest. Protected copyright. 43 44 disruptions in therapy. Curr Med Res Opin 2009;25:1247 60. 45 46 19. Colombo GL, Agabiti Rosei E, Margonato A, et al. Off patent generic medicines vs. off 47 48 patent brand medicines for six reference drugs: a retrospective claims data study from 49 50 51 five local healthcare units in the Lombardy Region of Italy. PLoS One 2013;8:e82990. 52 53 20. Hakonsen H, Toverud EL. Special challenges for drug adherence following generic 54 55 substitution in Pakistani immigrants living in Norway. Eur J Clin Pharmacol 56 57 2011;67:193 201. 58 59 22 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 23 of 35 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 21. Degli Esposti L, Sangiorgi D, Buda S, et al. Discontinuità terapeutica nei pazienti 5 6 sottoposti a sostituzione tra farmaci equivalenti. Evidenze dal “mondo reale”. 7 8 Supplemento a Politiche sanitarie 2015;16:3. 9 10 11 22. Agenzia Italiana del Farmaco (AIFA). Guideline for the classification and conduction of 12 13 the observational studies on medicines. 2010. Available from: 14 15 https://www.agenziafarmaco.gov.it/ricclin/sites/default/files/files_wysiwyg/files/CIRCUFor peer review only 16 17 LARS/Circular%2031st%20May%202010.pdf (accessed September 30, 2014). 18 19 20 23. Catalan VS, LeLorier J. Predictors of long term persistence on statins in a subsidized 21 22 clinical population. Value Health 2000;3:417 26. 23 24 24. Calderón Larrañaga A, Diaz E, Poblador Plou B, et al. Non adherence to 25 26 antihypertensive medication: The role of mental and physical comorbidity. Int J Cardiol 27 28 2016;207:310 6. 29 30 31 25. Steinman MA, Landefeld CS, Rosenthal GE, et al. Polypharmacy and prescribing quality 32 33 in older people. J Am Geriatr Soc 2006;54:1516 23.
34 http://bmjopen.bmj.com/ 35 26. Volpe M, Chin D, Paneni F. The challenge of polypharmacy in cardiovascular medicine. 36 37 Fundam Clin Pharmacol 2010;24:9 17. 38 39 40 27. Frankenstein L, Clark AL, Ribeiro JP. Influence of sex on treatment and outcome in 41
42 chronic heart failure. Cardiovasc Ther 2012;30:182 92. on September 24, 2021 by guest. Protected copyright. 43 44 28. Barsheshet A, Brenyo A, Goldenberg I, et al. Sex related differences in patients' 45 46 responses to heart failure therapy. Nat Rev Cardiol 2012;9:234 42. 47 48 29. Haynes RB, McDonald H, Garg AX, et al. Interventions for helping patients to follow 49 50 51 prescriptions for medications. Cochrane Database Syst Rev 2002;(2):CD000011. 52 53 30. Ghate SR, Biskupiak JE, Ye X, et al. Hemorrhagic and thrombotic events associated with 54 55 generic substitution of warfarin in patients with atrial fibrillation: a retrospective analysis. 56 57 Ann Pharmacother 2011;45:701 12. 58 59 23 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 24 of 35 BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 31. O'Leary A, Usher C, Lynch M, et al. Generic medicines and generic substitution: 5 6 contrasting perspectives of stakeholders in Ireland. BMC Res Notes 2015;8:790. 7 8 32. Toverud EL, Hartmann K, Hakonsen H. A Systematic Review of Physicians' and 9 10 11 Pharmacists' Perspectives on Generic Drug Use: What are the Global Challenges? Appl 12 13 Health Econ Health Policy 2015;13 Suppl 1:S35 45. 14 15 33. ThiebaudFor P, Patel BV, peer Nichol MB, etreview al. The effect of switching only on compliance and 16 17 persistence: the case of statin treatment. Am J Manag Care 2005;11:670 4. 18 19 20 34. Phillips B, Roberts C, Rudolph A, et al. Switching statins: the impact on patient 21 22 outcomes. Br J Cardiol 2007;14:280 5. 23 24 35. Liew D, Webb K, Meerding W J, et al. Potential cardiovascular consequences of 25 26 switching from atorvastatin to generic simvastatin in the Netherlands. Neth Heart J 27 28 2012;20:197 201. 29 30 31 36. Poluzzi E, Veronese G, Piccinni C, et al. Switching among equivalents in chronic 32 33 cardiovascular therapies: 'real world' data from Italy. Basic Clin Pharmacol Toxicol
34 http://bmjopen.bmj.com/ 35 2016;118:63 9. 36 37 37. Kesselheim AS, Misono AS, Shrank WH, et al. Variations in pill appearance of 38 39 40 antiepileptic drugs and the risk of nonadherence. JAMA Intern Med 2013;173:202 8. 41
42 38. Perreault S, Blais L, Lamarre D, et al. Persistence and determinants of statin therapy on September 24, 2021 by guest. Protected copyright. 43 44 among middle aged patients for primary and secondary prevention. Br J Clin Pharmacol 45 46 2005;59:564 573. 47 48 39. Shin JY, Choi NK, Jung SY,et al. Overlapping medication associated with healthcare 49 50 51 switching among Korean elderly diabetic patients. J Korean Med Sci 2011;26:1461 8. 52 53 40. Colombo GL, Agabiti Rosei E, Margonato A, et al. Impact of substitution among generic 54 55 drugs on persistence and adherence: A retrospective claims data study from 2 Local 56 57 Healthcare Units in the Lombardy Region of Italy. Atheroscler Suppl 2016;21:1 8. 58 59 24 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 25 of 35 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 4 41. Krueger K, Botermann L, Schorr SG, et al. Age related medication adherence in patients 5 6 with chronic heart failure: A systematic literature review. Int J Cardiol 2015;184:728 35. 7 8 42. Jin J, Sklar GE, Min Sen Oh V, et al. Factors affecting therapeutic compliance: A review 9 10 11 from the patient's perspective. Ther Clin Risk Manag 2008;4:269 86. 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33
34 http://bmjopen.bmj.com/ 35 36 37 38 39 40 41
42 on September 24, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 25 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 26 of 35 BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 STROBE Statement—Checklist of items that should be included in reports of cohort studies 4 5 Item No Page 6 Recommendation no. 7 8 Title and abstract 1 (a) Indicate the study’s design with a commonly used term in the title or the 9 abstract 10 11 The title on page 1 in the main manuscript states the study design: 12 1 13 “Therapy discontinuation or substitution in patients with cardiovascular 14 disease, switching among different products of the same off-patent active 15 Forsubstance: peer a “real-world” review retrospective cohort study”only 16 17 (b) Provide in the abstract an informative and balanced summary of what 18 was done and what was found 19 20 The Abstract in the main manuscript provides a balanced summary of what 21 was done and what was found: 2 22 23 ABSTRACT 24 25 Objective: The present study investigated the effects of switching to different 26 products of the same off-patent active substance (brand-name or generic) on 27 therapy discontinuation or substitution with another molecule of the same 28 class, in patients with cardiovascular disease treated with statins and 29 antihypertensives in clinical practice. 30 31 Design: A retrospective cohort study in a “real-world” setting. 32 Setting: Analysis of data performed by integrating administrative databases 33 that included a total of approximately two million health-assisted individuals 34 http://bmjopen.bmj.com/ from three Italian Local Health Units located in three different regions of 35 Italy. 36 37 Participants: All patients aged ≥18 years with at least one prescription of 38 simvastatin, ramipril or amlodipine in the period January 1st to December 39 31st 2010 were included and followed-up for two years. 40 41 Main outcome measures: Prescription refills occurring during follow-up were
42 evaluated. Frequency of discontinuation of therapy or substitution with on September 24, 2021 by guest. Protected copyright. 43 another molecule of the same class (for example, from simvastatin to a 44 different statin) during follow-up was identified. 45 46 Results: During follow-up, therapy discontinuation or substitution was found 47 to be more frequent in patients switching to a different product of the same 48 active substance compared with non-switching patients (11.5% vs 10.8% and 49 22.2% vs 20.8% [p=0.002], respectively, in the simvastatin group; 7.6% vs 50 7.8% and 19.2% vs 17.0% [p<0.001], respectively, in the ramipril group; 4.0% 51 vs 3.5% and 24.6% vs 22.7% [p<0.001], respectively, in the amlodipine 52 group). These findings were partially confirmed by multivariate analysis. 53 Conclusions: Switches among products of the same active substance are 54 quite common in patients with cardiovascular disease. Our study suggests 55 that switching may expose patients to a higher risk of therapy 56 discontinuation or substitution. 57 58 59 60 1 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 27 of 35 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 Introduction 4 5 Background/ 2 Explain the scientific background and rationale for the investigation being 6 reported 7 rationale 8 In the introduction in the main manuscript a scientific background and 9 rationale is provided: 4,5 10 11 A number of studies have demonstrated that patients often discontinue 12 long-term treatment or take less than prescribed, and that such non- 13 adherence reduces the potential preventive benefits.[3] 14 15 For peerMany reasons contribute review to patient non-adherence only to medical therapy, 16 such as ageing, comorbidities, poor relationship between patient and 17 physician, poor memory, and patients’ low perception of disease severity.[4] 18 In addition, the need to take several drugs concomitantly, or other 19 medication-related factors, may make remembering when to take each drug 20 more difficult and increase the risk of possible side effects caused by adverse 21 drug-drug interactions. Although medication side effects are probably not 22 the main cause of poor adherence, as there seems to be little direct 23 relationship between adherence and drug class,[3] they are also associated 24 with treatment discontinuation, especially in the early treatment of 25 hypertension [5-8]. Furthermore, some studies have demonstrated that 26 switching between different products of the same active substance can have 27 an impact on adherence to medication, because variation in packaging and 28 pill appearance may reduce adherence, especially for chronic diseases.[9, 10] 29 There is a perception among patients and physicians alike that frequent 30 changes between branded and unbranded products (as well as between 31 generics), all containing the same active substance, and especially if patients 32 are older and on multi-drug regimens, may cause patients to become anxious 33 when the appearance of their drugs changes.[11-13] This can lead to an
34 increased risk of patients making mistakes or double medicating, which flows http://bmjopen.bmj.com/ 35 on to increased drug non-adherence.[14, 15] 36 Few studies have investigated clinical differences related to switching among 37 different products of the same active substance in the cardiovascular setting. 38 Until now, most research has focused only on comparing brand-name and 39 generic drugs.[16-20] 40 41 Objectives 3 State specific objectives, including any prespecified hypotheses
42 on September 24, 2021 by guest. Protected copyright. 43 The objectives are stated in the introduction (paragraph 5) of the main 5 44 manuscript: 45 46 The aim of the present study was to investigate the effects of switching to 47 different products of the same off-patent active substance (brand-name or 48 generic) on therapy discontinuation or substitution with another molecule of 49 the same class, in patients with CVD treated with statins and 50 antihypertensives in clinical practice. 51 52 Methods 53 54 Study design 4 Present key elements of study design early in the paper 55 56 The key elements are presented in the method section of the main 57 manuscript: 58 59 60 2 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 28 of 35 BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 Data collection 4 5 The data used for the analysis were obtained from the administrative 6 databases of three Local Health Units (LHUs), whose databases included a 7 total population of about two million health-assisted individuals, in the 5,6 8 Italian regions of Lombardy, Lazio and Campania. We analyzed the following 9 archives: Assisted Subjects’ Database, containing the personal data of 10 patients; Medication Prescription database, containing all the information 11 relating to individual prescriptions, such as the International Nonproprietary 12 Names (INN) for pharmaceutical substances, the Anatomical-Therapeutic- 13 Chemical (ATC) code of the prescribed drug, the number of packages, the 14 number of units per package, the dose, the cost per unit and the date of the prescription; Hospital Discharge Database (SDO), containing information on 15 Foreach peer hospital discharge, review in particular the date only of admission and discharge, 16 primary and secondary diagnoses coded according to the International 17 Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM). 18 The patient code in each database allowed electronic linking among all 19 databases. To guarantee patient privacy, this patient code was transcoded 20 into an anonymous univocal numeric code. No identifiers related to patients 21 were provided to the researchers. According to the Italian law for 22 confidentiality of data[22] the study was notified to the Ethic Committees of 23 each LHU. 24 25 Cohort definition 26 27 The study was a retrospective cohort study including all patients aged ≥18 28 years that, between January 1st 2010 and December 31st 2010 (enrollment 29 period), had at least one prescription of simvastatin [ATC code: C10AA01], 30 ramipril [ATC code: C09AA05] or amlodipine [ATC code: C08CA01]. The date 31 of enrollment was defined as the earliest date within the enrollment period 32 in which the patient had the last switch of medication (switchers) or the last 33 prescription in the case of a patient continuing with the same medication
34 (non-switchers). Starting from this date, the individual patient was followed http://bmjopen.bmj.com/ 35 for two years (follow-up period). Switchers were defined as those patients 36 who switched among different products of the same off-patent active 37 substance (i.e., from brand-name to generic, from generic to brand-name, or 38 from generic to another generic). The changes in dose and dosage form were 39 not accounted for during switching. The definitions of the patient cohorts are 40 shown in Figure 1. Data on baseline characteristics, including demographics, 41 cardiovascular risk factors and polymedication, were collected; specifically, polymedication were identified as more than one cardiovascular medication 6,7 42 on September 24, 2021 by guest. Protected copyright. 43 at the date of enrollment. The data on drug prescriptions and 44 hospitalizations that occurred during the 12 months preceding the date of 45 enrollment were analyzed (characterization period). 46 Only patients with at least one prescription of the index drug in the previous 47 12 months were included (to capture patients who could have made a 48 change in therapy) and with at least two prescriptions at follow-up (to 49 include patients with continuity of treatment). Patients treated with fixed 50 combinations of the molecules under consideration (ramipril and diuretics 51 [ATC code: C09BA05], perindopril and amlodipine [ATC code: C09BB04], 52 ramipril and felodipine [ATC code: C09BB05], olmesartan and amlodipine 53 [ATC code: C09DB02], simvastatin and ezetimibe [ATC code: C10BA02]) were 54 excluded. 55 56 Patients transferred to another LHU during the follow-up period were 57 excluded from the analysis. 58 59 60 3 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 29 of 35 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 Study population 4 5 Cardiovascular risk 6 Patients were classified as high cardiovascular risk if they had cardiovascular 7 treatment or hospitalization for diabetes. Hospitalizations related to diabetes 8 were identified by ICD-9-CM code: 250 (primary discharge reasons); and/or 9 cardiovascular risk factors (previous hospitalization for Ischemic Heart 10 Disease [Acute myocardial infarction (ICD-9-CM: 410) Acute Cardiac Ischemia 11 (ICD-9-CM: 411), Angina Pectoris (ICD-9-CM: 413), Chronic Cardiac Ischemia 12 (ICD-9-CM: 414)]; Cerebral Hemorrhage [ICD-9-CM: 431]; Cerebral Artery 13 Occlusion [ICD-9-CM: 434]; Transient Cerebral Ischemia [ICD-9-CM: 435]; 14 Cerebral Circulatory Disorders [ICD-9-CM: 436]; Atherosclerosis [ICD-9-CM: 15 For440]; peer Other Peripheral review Vascular Disease [ICD-9-CM: only 443]; Chronic Renal 16 Failure [ICD-9-CM: 585]; Coronary Angioplasty [ICD-9-CM procedure: 0066, 17 360]); and/or at least two prescriptions of antidiabetic drugs [ATC code: 18 A10]. All other patients were classified as at moderate cardiovascular risk. 19 20 Drug treatments 21 Patients were also characterized by strategy of treatment at baseline with 22 23 lipid-lowering drugs (ATC code: C10) and antihypertensive drugs (ATC codes: 24 C02, C03, C07, C08, C09). 25 Data analysis at follow-up 26 27 During the follow-up period, the discontinuation or the first substitution of 28 therapy was identified. Discontinuation of therapy was defined as the 29 absence of prescriptions of the same therapeutic class (ATC group) as the 30 index molecule in the last quarter of observation. A substitution of therapy 31 was defined as a change to a different active substance of the same 32 therapeutic class (ATC group) (i.e., switching from simvastatin to a different 7,8 33 statin). A switch among different products of the same active substance was 34 identified by INN. http://bmjopen.bmj.com/ 35 36 Setting 5 Describe the setting, locations, and relevant dates, including periods of 5 37 recruitment, exposure, follow-up, and data collection 38 39 The setting, locations and relevant dates are presented in the method 40 section (paragraphs “Data collection”, “Cohort definition” and “Study 41 population”) of the main manuscript:
42 on September 24, 2021 by guest. Protected copyright. 43 The data used for the analysis were obtained from the administrative 44 databases of three Local Health Units (LHUs), whose databases included a 45 total population of about two million health-assisted individuals, in the 46 Italian regions of Lombardy, Lazio and Campania. 47 48 Participants 6 (a) Give the eligibility criteria, and the sources and methods of selection of NA 49 participants. Describe methods of follow-up 50 51 (b) For matched studies, give matching criteria and number of exposed and 52 unexposed 53 54 The participants are described in the methods section (paragraphs “Cohort 55 definition” and “Study population”) of the main manuscript: 56
57 The study was a retrospective cohort study including all patients aged ≥18 58 years that, between January 1st 2010 and December 31st 2010 (enrollment 59 60 4 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 30 of 35 BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 period), had at least one prescription of simvastatin [ATC code: C10AA01], 6 4 ramipril [ATC code: C09AA05] or amlodipine [ATC code: C08CA01]. (…) Only 5 patients with at least one prescription of the index drug in the previous 12 6 months were included (to capture patients who could have made a change in 7 therapy) and with at least two prescriptions at follow-up (to include patients 8 with continuity of treatment). Patients treated with fixed combinations of the 9 molecules under consideration (ramipril and diuretics [ATC code: C09BA05], 7 10 perindopril and amlodipine [ATC code: C09BB04], ramipril and felodipine 11 [ATC code: C09BB05], olmesartan and amlodipine [ATC code: C09DB02], 12 simvastatin and ezetimibe [ATC code: C10BA02]) were excluded.
13 Patients transferred to another LHU during the follow-up period were 14 excluded from the analysis. (…) Patients were classified as high 15 Forcardiovascular peer risk if reviewthey had cardiovascular treatmentonly or hospitalization for 16 diabetes. (…) All other patients were classified as at moderate cardiovascular 17 risk. 18
19 The methods of follow-up are described in the methods section 20 (paragraphs “Cohort definition” and “Study population”) of the main 21 manuscript: 22 23 (…) the individual patient was followed for two years (follow-up period). 24 7 25 Switchers were defined as those patients who switched among different products of the same off-patent active substance (i.e., from brand-name to 26 generic, from generic to brand-name, or from generic to another generic). 27 (…) During the follow-up period, the discontinuation or the first substitution 28 of therapy was identified. Discontinuation of therapy was defined as the 29 absence of prescriptions of the same therapeutic class (ATC group) as the 30 index molecule in the last quarter of observation. A substitution of therapy 31 was defined as a change to a different active substance of the same 32 therapeutic class (ATC group) (i.e., switching from simvastatin to a different 33 statin). A switch among different products of the same active substance was 34 http://bmjopen.bmj.com/ identified by INN. 35 36 37 38 6 39 40 41
42 on September 24, 2021 by guest. Protected copyright. 43 44 8 45 46 47
48 49
50 51
52 53 54 55 56 57 Variables 7 Clearly define all outcomes, exposures, predictors, potential confounders, 58 59 60 5 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 31 of 35 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 and effect modifiers. Give diagnostic criteria, if applicable 4 5 The outcomes are described in the methods section of the main 6 manuscript: 7 8 8 During the follow-up period, the discontinuation or the first substitution of 9 therapy was identified. Discontinuation of therapy was defined as the 10 absence of prescriptions of the same therapeutic class (ATC group) as the 11 index molecule in the last quarter of observation. A substitution of therapy 12 was defined as a change to a different active substance of the same 13 therapeutic class (ATC group) (i.e., switching from simvastatin to a different 14 statin). A switch among different products of the same active substance was 15 Foridentified peer by INN. review only 16 17 Data sources/ 8* For each variable of interest, give sources of data and details of methods of 18 measurement assessment (measurement). Describe comparability of assessment methods 19 if there is more than one group 20 21 Data sources and details of methods of assessment are given in the 22 methods section first paragraph:
23 24 The data used for the analysis were obtained from the administrative
25 databases of three Local Health Units (LHUs), whose databases included a total population of about two million health-assisted individuals, in the 26 5 27 Italian regions of Lombardy, Lazio and Campania. (...) The study was a retrospective cohort study including all patients aged ≥18 years that, 28 29 between January 1st 2010 and December 31st 2010 (enrollment period), had at least one prescription of simvastatin [ATC code: C10AA01], ramipril [ATC 30 31 code: C09AA05] or amlodipine [ATC code: C08CA01]. (…) the discontinuation 32 or the first substitution of therapy was identified. Discontinuation of therapy 33 was defined as the absence of prescriptions of the same therapeutic class
34 (ATC group) as the index molecule in the last quarter of observation. A http://bmjopen.bmj.com/ 35 substitution of therapy was defined as a change to a different active 36 substance of the same therapeutic class (ATC group) (i.e., switching from 37 simvastatin to a different statin). A switch among different products of the 38 same active substance was identified by INN. 8 39 40 41
42 Bias 9 Describe any efforts to address potential sources of bias NA on September 24, 2021 by guest. Protected copyright. 43 44 This is addressed by the design of the study and the clearly defined data 45 selection specifications 46 47 Study size 10 Explain how the study size was arrived at NA 48 49 How the study size was arrived at is described in the methods section 50 51 Sample size dictated by the source databases 52 53 Quantitative 11 Explain how quantitative variables were handled in the analyses. If 54 variables applicable, describe which groupings were chosen and why 55 56 The handling of the variables is described in the “Statistical analysis” 57 paragraph: 8 58 59 60 6 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 32 of 35 BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 Continuous variables were reported as mean ± standard deviation (SD) and 4 compared using the Student's t test; categorical variables were reported as 5 absolute numbers and percentages and compared using the chi-square test. 6 7 Discontinuations of therapy and substitution with another molecule of the 8 same class were analyzed by multivariate analysis using Cox proportional 9 hazards models; covariates considered in the models were: age, male sex, 10 high cardiovascular risk, cardiovascular treatments, change of formulation in 11 the period of characterization. 12 The analysis of Schoenfeld residuals (scaled and unscaled) was conducted to 13 assess the proportionality of risk. 14 15 Forp peer Values <0.05 were review considered statistically significant.only All analyses were 16 performed using STATA 12.0 SE. 17 18 Statistical 12 (a) Describe all statistical methods, including those used to control for 8 19 methods confounding 20 21 (b) Describe any methods used to examine subgroups and interactions 8 22 23 (c) Explain how missing data were addressed NA 24 25 (d) If applicable, explain how loss to follow-up was addressed NA 26 27 (e) Describe any sensitivity analyses 8 28 29 The statistical methods are given in the “Statistical analysis” paragraph: 30 31 Continuous variables were reported as mean ± standard deviation (SD) and 32 compared using the Student's t test; categorical variables were reported as 33 absolute numbers and percentages and compared using the chi-square test.
34 http://bmjopen.bmj.com/ 35 Discontinuations of therapy and substitution with another molecule of the 36 same class were analyzed by multivariate analysis using Cox proportional 37 hazards models; covariates considered in the models were: age, male sex, 38 high cardiovascular risk, cardiovascular treatments, change of formulation in 39 the period of characterization. 40 The analysis of Schoenfeld residuals (scaled and unscaled) was conducted to 41 assess the proportionality of risk.
42 on September 24, 2021 by guest. Protected copyright. 43 p Values <0.05 were considered statistically significant. All analyses were 44 performed using STATA 12.0 SE. 45 46 Results 47 48 Participants 13* (a) Report numbers of individuals at each stage of study—eg numbers 49 potentially eligible, examined for eligibility, confirmed eligible, included in 50 51 the study, completing follow-up, and analysed 52 The numbers of individuals are given in the Results section, paragraphs 53 “Simvastatin”, “Ramipril” and “Amlodipine” 54 9, 13, 55 56 (b) Give reasons for non-participation at each stage NA 57 58 59 60 7 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 33 of 35 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 (c) Consider use of a flow diagram NA 4 5 Descriptive data 14* (a) Give characteristics of study participants (eg demographic, clinical, social) 6 and information on exposures and potential confounders 7 8 The characteristics of study participants are given in Table 1. 9 9 10 11 (b) Indicate number of participants with missing data for each variable of NA 12 interest 13 14 (c) Summarise follow-up time (eg, average and total amount) 15 For peer review only 16 (…) followed for two years (follow-up period) 6 17 18 Outcome data 15* Report numbers of outcome events or summary measures over time 19 20 Numbers of outcome events were reported in the Results section, 8,9 21 paragraphs “Simvastatin”, “Ramipril” and “Amlodipine” and in Tables 2 10,11 22 and 3 23 24 Main results 16 (a) Give unadjusted estimates and, if applicable, confounder-adjusted 25 estimates and their precision (eg, 95% confidence interval). Make clear 26 which confounders were adjusted for and why they were included 27 28 Variables defined in Statistical analysis section and analyses reported in 29 Table 4 30 8 31 11,12 32 33 (b) Report category boundaries when continuous variables were categorized NA
34 http://bmjopen.bmj.com/ 35 (c) If relevant, consider translating estimates of relative risk into absolute NA 36 risk for a meaningful time period 37 38 Other analyses 17 Report other analyses done—eg analyses of subgroups and interactions, and 39 sensitivity analyses 40 41 Multivariate analysis is described in the “Statistical analysis” paragraph
42 and results are given in Table 4: 8, on September 24, 2021 by guest. Protected copyright. 43 11,12 44 Discontinuations of therapy and substitution with another molecule of the 45 same class were analyzed by multivariate analysis using Cox proportional 46 hazards models; covariates considered in the models were: age, male sex, 47 high cardiovascular risk, cardiovascular treatments, change of formulation in 48 the period of characterization. 49 50 Discussion 51 52 Key results 18 Summarise key results with reference to study objectives 53 54 Key results are summarized as follows in the Discussion: 16,17 55 56 (…) our study indicates that in a “real-world” setting, changes among 57 different product of the same active substance, including switching brand- 58 name to generic, generic to another generic and generic to brand-name, are 59 60 8 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 34 of 35 BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 quite common among patients with CVD. Our findings suggest that switching 4 to a different product of the same off-patent active substance, brand-name 5 or generic, may expose patients to a higher risk of therapy discontinuation or 6 substitution than continuing treatment with the same product. 7 8 Limitations 19 Discuss limitations of the study, taking into account sources of potential bias 9 or imprecision. Discuss both direction and magnitude of any potential bias 10 11 Limitations are summarized as follows in the Discussion: 12 16 13 Our analysis has several limitations inherent to any observational study. 14 First, the study was performed using the administrative databases, and the 15 Forreasons peer for switch, non-adherencereview or discontinuation only of treatment in the 16 patients were not retrievable from the dataset. Also, no information on the 17 role of the counseling pharmacist when substituting and dispensing drug 18 packages were available to us. A second limitation is a relatively limited 19 sample size. Although in our study we used the healthcare databases of 20 Lombardy, Lazio and Campania, three Italian Regions localized from north to 21 south of Italy, including data for a total population of about 2 million, and 22 considering that we have focused our analysis among users of simvastatin, 23 ramipril, and amlodipine, larger studies are needed to confirm and to enhance the generalizability of the findings, and in different populations. 24 25 Interpretation 20 Give a cautious overall interpretation of results considering objectives, 26 limitations, multiplicity of analyses, results from similar studies, and other 27 relevant evidence 28 29 The interpretation is given as follows in the Discussion: 30 31 In accordance with previous studies,[23-26] this retrospective analysis in a 14-16 32 “real-world” setting shows that age, gender, cardiovascular risk, and 33 polypharmacy affect discontinuation of therapy. (…) The data of this study 34 show also that for the same active substance, a change of product http://bmjopen.bmj.com/ 35 (regardless of whether it is a brand-name or generic drug) increases the risk 36 of discontinuation of therapy and of substitution with another molecule of 37 the same class. (…) we cannot exclude the possibility that other potential 38 determinants can play a key role in a reduction of patients’ adherence. 39 Observational studies [33, 34] have demonstrated a relationship between 40 age, gender, cardiovascular risk factors, polymedication and a suboptimal 41 adherence to therapy; our findings are in agreement with these previous
42 analyses. In contrast, new evidence suggested that older age is not related to on September 24, 2021 by guest. Protected copyright. 43 poorer medication adherence to cardiovascular medication. A recent 44 systematic search of the bibliographic database MedLine and all Cochrane 45 databases, analyzing the relationship between age and medication 46 adherence in adult patients with chronic heart failure (CHF), showed that 47 that older age alone is not related to poorer medication adherence 48 compared with younger patients with CHF.[35] Our study does not support 49 this concern. 50 However, the underlying reasons for poor adherence are not fully 51 understood, and there may be many reasons behind these behaviours, some 52 of which relate to the perceptions that physicians, pharmacists and patients 53 may have of drugs and therapies. Nevertheless, the clinical consequences 54 that may result from these perceptions are important and should be 55 considered. 56 57 Generalisability 21 Discuss the generalisability (external validity) of the study results 58 59 60 9 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 35 of 35 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-012003 on 2 November 2016. Downloaded from
1 2 3 The issue of generalizability is addressed as follows in the Discussion: 16,17 4 5 Although in our study we used the healthcare databases of Lombardy, Lazio 6 and Campania, three Italian Regions localized from north to south of Italy, 7 including data for a total population of about 2 million, and considering that 8 we have focused our analysis among users of simvastatin, ramipril, and 9 amlodipine, larger studies are needed to confirm and to enhance the 10 generalizability of the findings, and in different populations. 11 12 Despite these limitations, our study indicates that in a “real-world” setting, changes among different product of the same active substance, including 13 switching brand-name to generic, generic to another generic and generic to 14 brand-name, are quite common among patients with CVD. Our findings 15 Forsuggest peer that switching review to a different product ofonly the same off-patent active 16 substance, brand-name or generic, may expose patients to a higher risk of 17 therapy discontinuation or substitution than continuing treatment with the 18 same product. 19 20 21 Other information 22 23 Funding 22 Give the source of funding and the role of the funders for the present study 24 and, if applicable, for the original study on which the present article is based 25 26 Such information is given in the Acknowledgments: 18 27 28 Funding: 29 This research received no specific grant from any funding agency in the 30 public, commercial or not-for-profit sectors. 31 32 *Give information separately for exposed and unexposed groups. 33
34 http://bmjopen.bmj.com/ 35 36 Note: An Explanation and Elaboration article discusses each checklist item and gives methodological background and 37 published examples of transparent reporting. The STROBE checklist is best used in conjunction with this article (freely 38 available on the Web sites of PLoS Medicine at http://www.plosmedicine.org/, Annals of Internal Medicine at 39 http://www.annals.org/, and Epidemiology at http://www.epidem.com/). Information on the STROBE Initiative is 40 41 available at http://www.strobe-statement.org.
42 on September 24, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 10 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml