Journal of the New Zealand Medical Association Vol 128 | No 1416 | 12 June 2015
Barriers to successful cessation among young late-onset smokers
An important investment to control Acute Rheumatic Fever needs to run its course
Whooping Karyotypes, confined blood HIV-associated tuberculosis in chimerism, and confusion: a Auckland cough—where case of genetic sex mislabelling Direct access GP referral for are we now? and its potential consequences ETT functions as a virtual clinic Publication Information published by the New Zealand Medical Association
NZMA Chairman To contribute to the NZMJ, first read: Dr Stephen Child www.nzma.org.nz/journal/contribute
NZMJ Editor Other enquiries to: Professor Frank Frizelle NZMA PO Box 156 NZMA Communications Manager The Terrace Sharon Cuzens Wellington 6140 Phone: (04) 472 4741 NZMJ Production Jeremiah Boniface © NZMA 2015
To subscribe to the NZMJ, email [email protected]
Subscription to the New Zealand Medical Journal is free and automatic to NZMA members. Private subscription is available to institutions, to people who are not medical practitioners, and to medical practitioners who live outside New Zealand. Subscription rates are below. All access to the NZMJ is by login and password, but IP access is available to some subscribers. Read our Conditions of access for subscribers for further information www.nzma.org.nz/journal/subscribe/conditions-of-access If you are a member or a subscriber and have not yet received your login and password, or wish to receive email alerts, please email: [email protected] The NZMA also publishes the NZMJ Digest. This online magazine is sent out to members and subscribers 10 times a year and contains selected material from the NZMJ, along with all obituaries, summaries of all articles, and other NZMA and health sector news and information.
Subscription rates for 2015 New Zealand subscription rates Overseas subscription rates Individuals* $290 Individual $402 Individual article $25 Institutions $543 Individual article $25
*NZ individual subscribers must not be doctors (access is via NZMA Membership) New Zealand rates include GST. No GST is included in international rates. Note, subscription for part of a year is available at pro rata rates. Please email [email protected] for more information. Individual articles are available for purchase by emailing [email protected]
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 2 www.nzma.org.nz/journal CONTENTS
EDITORIAL 62 6 Karyotypes, confined blood An important investment to control chimerism, and confusion: a case Acute Rheumatic Fever needs to run of genetic sex mislabelling and its its course potential consequences Diana Lennon, Joanna Stewart Aarthi Ravishankar, José G B Derraik, Sarah Mathai, Wayne S Cutfield, Paul L Hofman ARTICLES VIEWPOINT 10 Ethnic differences in acute 66 hospitalisations for otitis media Domperidone safety: a mini-review and elective hospitalisations for of the science of QT prolongation and ventilation tubes in New Zealand clinical implications of recent global children aged 0–14 years regulatory recommendations Justine McCallum, Liz Craig, Ian Whittaker, Pamela J Buffery, R. Matthew Strother Joanne Baxter CLINICAL CORRESPONDENCE 21 75 Whooping cough—where are we now? Not a simple back pain A review Jen-Li Looi, Ruvin Gabriel Tomasz Kiedrzynski, Ange Bissielo, Mishra Suryaprakash, Don Bandaranayake 77 Recalcitrant peripheral spondyloarthritis 28 treated with radiotherapy The role of echocardiography in Nur Azri Bin Haji Mohd Yasin, Shaun Anthony Staphylococcus aureus bacteraemia at Costello Auckland City Hospital Nicholas Gow, Boris S. Lowe, LETTERS Joshua Freeman, Sally Roberts 79 36 Direct access GP referral for ETT HIV-associated tuberculosis functions as a virtual clinic in Auckland Jessica H Greaves, Joan D Leighton, John G Christopher Luey, David Milne, Simon Briggs, Lainchbury, Paul G Bridgman Mark Thomas, Rupert Handy, Mitzi Nisbet. 81 44 Uptake of new medicines: the Efficacy of intralesional triamcinolone Pharmaceutical Management Agency injections for benign refractory of New Zealand (PHARMAC) in the oesophageal strictures at Counties international context Manukau Health, New Zealand Rajan Ragupathy, Zaheer-Ud-Din Babar Yeri Ahn, Christin Coomarasamy, Ravinder Ogra 84 Improving our strategy to prevent and 51 control measles outbreaks Barriers to successful cessation Lance Gravatt among young late-onset smokers Hayley Guiney, Judy Li, Darren Walton 87 Methuselah 88 100 Years Ago: The War, the Birth Rate, and Strong Drink
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 3 www.nzma.org.nz/journal SUMMARIES
An important investment to control Acute Rheumatic Fever needs to run its course Diana Lennon, Joanna Stewart Rheumatic Fever reduces life spans by 10 years, but is preventable. Current school programmes to prevent it in high-risk areas are promising, but numbers are needed to prove it is working to prevent rheumatic fever. The Ministry of Health has co-funded the school programmes with DHBs, but is handing over to the DHBs before there is proof that the programme is effective (numbers need to accrue to prove statistical significance). DHBs may not continue to fund and have no evidence to inform their funding choices. This is a short-term wasted investment.
Whooping cough—where are we now? Tomasz Kiedrzynski, Ange Bissielo, Mishra Suryaprakash, Don Bandaranayake This paper describes the recent trends of pertussis and vaccine uptake in New Zealand based on notifications and immunisation registration information since 2011. It highlights the current risk for the infant in the first months after birth and the crucial role a pertussis booster in pregnancy could play. It also aims to show that protection of infants by the current vaccine can be improved by timely immunisation even in a situation of improving overall uptake rates that are nearing the national target of 95%.
The role of echocardiography in Staphylococcus aureus bacteraemia at Auckland City Hospital Nicholas Gow, Boris S. Lowe, Joshua Freeman, Sally Roberts Staphylococcus aureus blood stream infection (SAB) is a major cause of morbidity and mortality in New Zealand. When it is diagnosed, it is important to exclude endocarditis (infection affecting predominantly the heart valves) as this condition is associated with a 20% risk of death. Therefore international and local guidelines suggest all cases of SAB should be investigated with a cardiac ultrasound study known as an echocardiogram. This study of the use of echocardiography in SAB at Auckland City Hospital has revealed that some patients are at particularly low risk of endocarditis; in this group it may be possible to avoid doing an echocardiogram, therefore making more valuable use of this resource.
Efficacy of intralesional triamcinolone injections for benign refractory oesophageal strictures at Counties Manukau Health, New Zealand Yeri Ahn, Christin Coomarasamy, Ravinder Ogra Some benign strictures in the food pipe can be very difficult to manage and require endoscopy and stretching every few weeks and can severely affect the quality of life of such patients. This paper describes a simple steroid injection into the scar tissue at the time of endoscopy that reduces the number of procedures and allows patients to swallow better.
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 4 www.nzma.org.nz/journal SUMMARIES
Karyotypes, confined blood chimerism, and confusion: a case of genetic sex mislabelling and its potential consequences Aarthi Ravishankar, José G B Derraik, Sarah Mathai, Wayne S Cutfield, Paul L Hofman Disorders of sex development (DSD) encompass a range of conditions which at their extreme can present as boys appearing like normal girls and girls appearing like normal boys. The management of infants with DSD can be extremely complex, requiring the long-term involvement of a multidisciplinary team working alongside the family. However, in some cases a normal infant may be wrongly labelled as having DSD, which may lead to irreversible consequences such as inappropriate surgery. This case illustrates how easily this can occur. The infant’s blood had a normal male 46 XY karyotype due to bone marrow transplant from the male twin while the other tissues were a normal female 46XX karyotype. This led to an incorrect sex determination at birth and almost to the potentially disastrous surgical removal of normal ovaries.
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 5 www.nzma.org.nz/journal EDITORIAL
An important investment to control Acute Rheumatic Fever needs to run its course Diana Lennon, Joanna Stewart
ew Zealand has been applauded The age group at greatest risk for ARF is publically at the World Health primary-school-aged children (range 5–12 NAssembly, Geneva for the commit- years) and is the best group for investment ment of the New Zealand Government in an effective prevention programme. moving towards control of Acute Rheumatic The government has wisely focused on Fever (ARF).1 Our primary prevention strat- this age group, investing in an innovative egy for control of first episodes of ARF is primary health care delivery model through innovative. Other countries seek guidance schools. Nurses, under delegated authority from these programmes. This investment for treatments, work in partnership with is an important step forward for the health community health workers, with the of New Zealand children and adults. The potential to prevent the most ARF cases life long shadow of the sequela of ARF, per health dollar invested. School clinics Rheumatic Heart Disease (RHD) can reduce are operating in focused areas of high ARF longevity by up to 10 years.2 endemicity throughout the North Island. ARF is an important child health indicator Many studies have highlighted the poor in New Zealand with wide socio-economic access for children in low socio-economic and ethnic disparities. Poor children groups, particularly Māori and Pasifika residing in decile 10 (using the New Zealand children.7 Children in families with limited Deprivation Score) have been shown to be transport and time during surgery hours nearly 30 times more likely to have ARF are less likely to access medical care. This compared to those socioeconomically better approach to improved access to health care off.3 A child living in the poorest area, through schools was studied in a cluster decile 10, has a 1 in 150 cumulative risk randomised Health Research Council/Heart of being hospitalised for ARF by his or her Foundation/Ministry of Health funded 15th birthday.4 This disparity is likely due to controlled trial in New Zealand schools over multiple factors eg: household crowding for 4 years (n= ~25 000 subjects) in an area of a very infectious disease (group A strepto- endemic ARF and provided evidence for coccal (GAS) pharyngitis as the precursor); the New Zealand scene.8 A meta-analysis families unaware ARF is a preventable of community- and school-based studies, disease; and poor access to primary including the New Zealand study, suggested healthcare. Most importantly, the strongest ~60% of ARF cases could be prevented published evidence for control comes from by adequate access to primary care and improving healthcare access with associated treatment of GAS.9 public health messaging.5 Other investments by the government GAS pharyngitis, the preceding trigger for to widen the approach to control of ARF ARF is a very common childhood illness. by sore throat management are laudable, Important considerations are the frequency eg, improved free access at GP practices and the associated relatively minor in areas at high risk of ARF. However, this symptoms (with a short natural history of effect is unproven and so far the uptake symptoms and signs without treatment of 1-3 is difficult to evaluate. The influence of days) of a GAS pharyngitis.6 Persistence of extending free health care to children <13 the GAS has been shown to be an important years of age has yet to be seen. High school risk factor for emergence of ARF.6 clinics with free swabbing and treatment,
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 6 www.nzma.org.nz/journal EDITORIAL
recently put in place, are an important However improving access to primary extension of the school clinic evidence-sup- health care through school health services ported model.8 Acute first episode ARF in can only function 40 weeks of the year. adolescents is less common but still occurs.8 In the school study we attempted to In the New Zealand school-based study, high address this with free ‘drop-in’ clinics schools were part of the design. We found, in the school holidays at a centralised as many before us have, that influencing school or other venue without success. We adolescent health-seeking behaviour is a came to understand that the limitations different challenge from primary health (not including cost) of the traditional care delivery in primary schools, with primary health care model, such as general fewer students participating in the sore practice, are factors such as the limitations throat clinic process. Messages to improve of transport, hours of opening of the knowledge within high school populations facility, the ongoing need for reinforced were shown to be effective.10 A further ‘sore throats matter’ messages, and the study confirmed this in primary schools.11 limited ability of typically large households with many competing and more pressing The inability of the New Zealand School concerns. Bringing young children to randomised controlled trial (RCT) to medical attention for a seemingly harmless reach statistical significance seems likely and self-limited affliction becomes a to have been influenced by the inability, lower priority. Improving community as a result of the design, to follow up knowledge will address only some of these household members of a GAS pharyngitis concerns. However only recently have 12 index case as recommended. The inci- the recommendations to ensure improved dence of pharyngeal GAS in the RCT was knowledge of ARF prevention made by unchanged over the study period. The infec- the Heart Foundation-sponsored Working tiousness of pharyngeal GAS in a household Group16 enlarged to a focus on Pacific RF where nearly half of siblings will have a prevention messages. culture positive GAS pharyngitis is well To move towards holistic care and to described.13 This aspect has been corrected avoid a health care intervention focusing in the current school clinic model of GAS on a sole disease, an intervention was pharyngitis management where symp- piloted to treat the precursors of the most tomatic siblings are sought out for throat numerous cause of medical hospitalisations, culture and treatment.11 Its seems likely serious skin infection in the primary school that this measure will have an effect on age group. Using the school clinic model the household and community burden of developed for ARF prevention, nurse-led pharyngeal GAS. diagnosis and treatment of skin infection Measurement of cross sectional with standing orders for antimicrobial prevalence of pharyngeal GAS annually treatments if required (the minority) was in selected school-aged populations in evolved.11,17 This is now a part of the health different DHBs is underway, measuring care package through year 1–8 schools the necessary precursor of ARF (funded by in the Auckland region and many other HRC and partners). Preliminary findings North Island DHBs. As this is a numeri- show a reduction in the pharyngeal cally common cause of hospitalisation in GAS burden in high-risk communities the school-age group an effect might be since the commencement of the school available soon. This approach will consol- programmes, and an observed low idate the delivery of primary care services pharyngeal GAS burden in a population for better access for families. Certain condi- at low risk of ARF.14 An effect on ARF tions trigger an immediate referral to the rates will be statistically possible in family GP or hospital eg, orbital cellulitis. 2016/2017. An adaption of the school Ongoing sustainability of school sore model to a semi-rural environment in a throat management programmes in ARF small, predominately Māori community endemic areas seems obvious, at least until has proved successful in preventing ARF the investment has been robustly evaluated. over a >10-year period, and reduced Household conditions (crowding facilitates pharyngeal GAS in another area.14,15 GAS spread) are unlikely to change fast.
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 7 www.nzma.org.nz/journal EDITORIAL
The paradox is that Ministry of Health will not be available until 2016/2017. The officials have advised the government to MOH has indicated that funding for school fund this investment only until the end of clinics for ARF prevention will cease at the 2015–16 financial year. After this time the end of the 2015–16 financial year. The DHBs DHBs concerned will make internal deci- will be without evidence to guide their sions on the continuation of the school decision making for ongoing investment in clinic programmes. Given the current this project. This is out of keeping with the financial constraints faced by DHBs, Finance Minister’s public declaration at the continued funding of ARF prevention Auckland University of Technology in 2013 programmes may not be prioritised, as that health funding decisions should be there is no policy in New Zealand for evidence-based.19 Continuation of funding consideration of children first. (and direction) from the MOH seems indis- Such an important investment to control putable until evidence has accrued for or new cases of ARF requires the support of against the current programme to attempt the best possible data to evaluate its effect.18 to prevent this relatively uncommon but Sufficient power to show a meaningful devastating disease. It would enable us to reduction of ARF as a result of the school hold our heads up the world arena for this programme (as part of the HRC evaluation) highly predictive child health indicator.
Competing interests: Funding for rheumatic fever research from 1997 has been received from the HRC, Heart Foundation of NZ and the Ministry of Health (and formerly North Health) Author information: Diana Lennon, Professor of Population Child & Youth Health, FMHS, Department of Paediatrics, University of Auckland, Auckland; Joanna Stewart, FMHS, Department of Epidemiology & Biostatistics, University of Auckland, Auckland. Corresponding author: Professor Diana Lennon, School of Population Health, University of Auckland, Tamaki Campus, 261 Morrin Road, Glen Innes, Private Bag 92019, Auckland [email protected] URL: http://www.nzma.org.nz/journal/read-the-journal/all-issues/2010-2019/ 2015/vol-128-no-1416/6557
REFERENCES: 1. http://www.who.int/ Youth Epidemiology Health. 2010;46(9):534-48. mediacentre/events/ Service; 2007. 8. Lennon D, Stewart J, Farrell governance/wha/en/ 4. Milne RJ, Lennon DR, E, Palmer A, Mason H. 2. Milne RJ, Lennon D, Stewart JM, Vander Hoorn School-Based Prevention of Stewart JM, Vander Hoorn S, Scuffham PA. Incidence Acute Rheumatic Fever: A S, Scuffham PA. Mortality of acute rheumatic fever Group Randomized Trial in and hospitalisation costs in New Zealand children New Zealand. Pediatr Infect of rheumatic fever and and youth. Journal of Dis J. 2009;28(9):787-94. rheumatic heart disease Paediatrics and Child 9. Lennon D, Kerdemelidis Health.2012;48 (8):685-91. in New Zealand. Journal M, Arroll B. Meta-Analysis of Paediatrics and Child 5. Gordis L. Effectiveness of Trials of Streptococcal Health. 2012;48 (8):692-7. of comprehensive-care Throat Treatment Programs 3. Craig E, Jackson C, Han programs in preventing to Prevent Rheumatic DY, NZCYES Steering rheumatic fever. N Engl J Fever. Pediatr Infect Dis Committee. Monitoring Med. 1973;289(7):331-5. J. 2009;28(7):e259-e64. the health of New Zealand 6. www.heartfoundation.org.nz 10. Harre N, Thomas D, Brown children and young people: 7. Kerdemelidis M, Lennon K, Raza F, Lennon D. indicator handbook. DR, Arroll B, Peat B, Jarman Communicating informa- Auckland: Paediatric J. The primary prevention tion about sore throats Society of New Zealand, of rheumatic fever. Journal and rheumatic fever to New Zealand Child and of Paediatrics and Child South Auckland high-
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 8 www.nzma.org.nz/journal EDITORIAL
school students. N Z Med in Australia. Pediatrics. 17. Vogel A, Lennon D, Gray J. 2000;113(1111):215-7. 2007;120(5):950-7. S, Farrell E, Anderson P. 11. Gray S, Lennon D, Ander- 14. Lennon D on behalf of : Ball Registered nurse assess- son P, Stewart J, Farrell E. S MJ, Farrell E, Anderson ment and treatment of Nurse-led school-based clin- P3, Percival T5, Crengle skin sepsis in New Zealand ics forskin infections and S5, Leversha A4, Carter P6, schools: the development rheumatic fever prevention Stewart J5. First Case Acute of protocols. N Z Med J. : results from a pilot study Rheumatic Fever (Arf) Is 2013;126(1380):27-38. in South Auckland. N Z Preventable: The Accruing 18. Moxon T, Lennon DR, Reed MedJ. 211 2013;126(1373). New Zealand Evidence. P, Jackson C, Anderson 12. American Academy of ASID conference 2015. P, Leversha A, et al. Is a Pediatrics Red Book. Group 15. Jarman J. Reducing Rheumatic Fever Register A streptococcal infections a striking health the best way to evaluate the - Red Book. 28th ed., Red inequality. N Z Med J. Government programme to Book: . Pickering LK, Baker 2011;124(1334):18-20. control Rheumatic Fever? (Abstract) Annual Scien- CJ, Kimberlin DW, Long 16. Lennon D, Convenor. tific Meeting Paediatric SS, editors. Illinois: ELK Advice for the Ministry of Society of New Zealand, Grove Village; 2009. Health for Best Practice for Palmerston North.2012. 13. Danchin MH, Rogers S, Rheumatic Fever Control. Kelpie L, Selvaraj G, Curtis Summary of International 19. Integrated Data Road N, Carlin JB, et al. Burden Workshop on Rheumatic Show at Auckland of acute sore throat and Fever/Rheumatic Heart University of Technology group A streptococcal Disease Control in New on 2 April, 2014. pharyngitis in school-aged Zealand. 2010. www. children and their families paediatrics.org.nz.
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 9 www.nzma.org.nz/journal ARTICLE
Ethnic differences in acute hospitalisations for otitis media and elective hospitalisations for ventilation tubes in New Zealand children aged 0–14 years Justine McCallum, Liz Craig, Ian Whittaker, Joanne Baxter
ABSTRACT AIMS: This paper describes ethnic differences in acute hospitalisations for otitis media (OM) and elective hospitalisations for ventilation tube insertion in New Zealand children aged 0–14 years. Ethnic differences in first attendances at Ear Nose and Throat (ENT) outpatient clinics are also described. METHOD: The analysis included all hospital admissions of children aged 0–14 years during 2002–2008 which met the following criteria: Acute admissions with an ICD–10-AM primary diagnosis code of otitis media; and elective admissions with a primary procedure code of ventilation tube insertion. First attendances at ENT outpatient clinics during 2007–2008 were also reviewed. Explanatory variables included ethnicity, gender, age, and NZ Deprivation Index decile. RESULTS: Among 0–4 year olds, Māori and Pacific children were more likely to be admitted acutely for otitis media than European children. In contrast, both Māori and Pacific children had lower rates of elective admissions for ventilation tube insertion, with ethnic differences being most marked for children from the most deprived areas. Māori and Pacific children aged 5–14 years also had higher acute otitis media admis- sion rates than European children. In contrast to their younger counterparts however, they also had higher rates of ventilation tube insertion. Exploration of ENT outpatient data for children 0–4 years revealed similar first appointment rates for European and Māori children, but lower rates for Pacific and Asian children. For the 5–14 age group, first appointment rates were higher for Māori and Pacific children than for European children. However, Māori and Pacific children in both age groups had higher rates of non-attendance at their first ENT appointments than European children. CONCLUSION: This study highlights ethnic differences in access to ventilation tubes amongst New Zealand’s 0–4 year olds, with the greatest inequalities being seen for Māori, Pacific and Asian children living in the most deprived areas. For Māori and Pacific children, such differences cannot be attributed to lower rates of AOM or OME compared to European children. The fact that similar patterns are not seen for children aged 5–14 years potentially suggests that routine Well Child hearing screening may be playing a role in identifying unmet need in this older age group. Such disparities also suggest that factors over and above OM prevalence may be influencing access to ventilation tubes. Further research is required to determine why Māori and Pacific children (0–4 years) have similar/lower ENT appointment rates than European children, despite a higher burden of middle ear disease, as well as higher non-attendance rates at outpatient clinics. Given the importance of early detection and treatment of OM for children’s ongoing well-being and educa- tion, a greater understanding of the reasons for these inequalities is urgently required.
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 10 www.nzma.org.nz/journal ARTICLE
found in 2008 that although Māori and Introduction Pacific children aged 0–4 years had higher By the age of 3 years the majority of acute hospital admission rates for OM than children (~80%) will have had one episode European children, they paradoxically of acute otitis media (AOM).1 Approximately had lower elective admission rates for the 10–30 % will have a further episode,2 and insertion of ventilation tubes.9 This paradox up to 25% will develop persistent otitis raises important questions regarding the media with effusion (OME) following AOM. nature of and mechanisms leading to such A common treatment of recurrent AOM disparities. At face value, they suggest that and persistent OME is the insertion of pre-school aged Māori and Pacific children ventilation tubes, more commonly referred have higher levels of need, yet may receive to as grommets. Ventilation tubes decrease lower levels of treatment. the number of episodes of AOM3 and can This study investigates ethnic differences improve hearing at 6 and 12 months after in acute hospital admissions for OM and insertion.4 elective admissions for ventilation tubes In New Zealand the general practi- in New Zealand children. It also explores tioner (GP) is the main source of referral access to first ENT Outpatient appoint- to specialist services including the Ear, ments by ethnicity, as ENT Outpatient Nose and Throat clinic (ENT) for consider- appointments are one of the key steps on ation for ventilation tubes. A child can be the pathway leading to ventilation tube referred to the GP and subsequently the insertion for New Zealand children. ENT clinic through a number of mecha- nisms. Prior to 2008, Well Child services Method included a specific hearing screening This study has two parts. The first program which aimed to identify children comprises an analysis of acute hospital failing tympanometry at 3 years of age admissions for otitis media and elective or screening audiometry at school entry admissions for ventilation tubes in children (5 years of age).5 This was replaced by aged 0–14 years. The second explores first the B4 School Check in 2008, which now outpatient appointments and attendances at comprises audiometry, followed by tympa- ENT clinics in the same age group. nometry (if necessary) at age 4, as part of the Well Child Tamariki Ora Programme.6 Hospitalisation data from the National Children can also be referred to GPs by Minimum Dataset (NMDS) was used to Well Child Providers who identify hearing calculate acute admission rates for otitis concerns during routine checks, with media and elective admission rates for children also presenting to GPs as a result ventilation tubes. The NMDS dataset of parental concerns, or more recently, contains mainly publicly funded hospital newborn hearing screening.(7,8) Following admissions, as privately funded private GP assessment, a child may be referred hospital data is incomplete. to an otolaryngologist, with the otolaryn- The analysis included all hospital admis- gologist then determining whether the sions of children aged 0–14 years during insertion of ventilation tubes is necessary. 2002–2008 which met the following criteria: The New Zealand Hearing and Screening 1. Acute admissions for otitis media Vision Report 2005–2006,5 which reports included all unplanned admissions on data from the earlier hearing screening (admitted on the day of presentation) programme described above, found that with an ICD10-AM primary diagnosis a higher proportion of Pacific and Māori code of H65, H66 or H67 (Otitis Media). children failed tympanometry screening 2. Elective admissions for venti- for OME at 3 years, than did European/ lation tubes included all Arranged Pākehā children. Pacific and Māori children (admission less than seven days after also had higher audiometry and/or tympa- a specialist decision) and Waiting nometry failure rates at school entry than List (admission seven or more days European/Pākehā children. after specialist decision) admissions Similarly, the New Zealand Child and with an Australian Classification Youth Epidemiology Service (NZCYES) of Health Interventions (ACHI)
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 11 www.nzma.org.nz/journal ARTICLE
primary procedure code of 4163200 presented as a decile scale with decile 1 or 4163201 (unilateral or bilateral representing the 10% least deprived small myringotomy with insertion of tube). areas, and decile 10 representing the 10% All children aged under 15 years were most deprived small areas. included, with children further being subdi- The second part of this study examined vided into two age groups (0–4 years and first appointment and attendance rates 5–14 years). These age groups were chosen at ENT outpatients, using data from the because they best reflected service delivery National Non–Admitted Patient Collection configurations: infants and preschool (NNPAC). Public outpatient clinic data children, covered by Well Child hearing was analysed for 2007–2008. The dataset surveillance and screening; and school age contains no diagnostic coding, such as ICD children, primarily cared for by GPs. codes, so analysis was by health specialty Analysis by other known risk factors for code only. middle ear disease was also undertaken. All children aged 0–14 years who had a Variables included age (known higher first outpatient appointment with an ENT prevalence in younger children),10 gender health specialty code (S25, S26 or S28) were (higher prevalence in males),11 ethnicity included. Those with follow up appoint- (higher prevalence in Indigenous popula- ments were excluded. All children with an tions)12 and NZ Deprivation Index quintile Attendance Code of ‘ATT’ were considered (known higher prevalence in children from to have attended their appointment, while areas of high deprivation).12 those with a ‘DNA/DNW’ code were recorded The denominator for the rate calcula- as either not attending or not waiting to be tions was the 2001 and 2006 New Zealand seen. The attendance rate was calculated Censuses (estimated resident population), by dividing the number attending (ATT) with linear extrapolation being applied by the total number of first appointments. between Census years. Ethnic specific rates Ethnic specific first appointment rates were were calculated by dividing the number calculated by dividing the number of first of hospital admission or outpatient events appointments by the census denominators for children of each ethnic group, by their for that ethnic group. relevant ethnic specific census denomi- nators. The rationale for the presentation of Results crude ethnic specific rates and rate ratios, Table 1 presents acute admissions for otitis rather than those which had been adjusted media in New Zealand children during 2002– for NZ Deprivation Index decile, was the 2008. During this period Māori and Pacific desire to reflect the actual inequalities expe- children aged 0–4 years had significantly rienced by Māori and Pacific children. higher acute admission rates for otitis media Level 1 prioritised ethnicity was used than those of European ethnicity. Rates for for both admission and outpatient numer- Māori children were 1.49 (95% CI 1.38–1.60) ators and for the denominator. Prioritised times higher than for European children, ethnicity means that for those children for while rates for Pacific children were 1.53 whom multiple ethnic affiliations are iden- (95% CI 1.38–1.69) times higher. Admission tified, a single ethnic group is assigned. This rates were also 1.25 (95% CI 1.17–1.33) times uses a Statistics New Zealand algorithm that higher for males than for females. results in people who report being Māori In the 5–14 year old age group, while as one or more of their ethnic groups being overall admission rates were much lower, assigned Māori ethnicity as a priority. This is Māori and Pacific children still had signifi- followed by prioritisation of Pacific ethnicity. cantly higher acute admission rates than did The child’s domicile code (usual area of European children. Rates for Māori children residence) was mapped to the New Zealand were 1.95 (95% CI 1.60–2.37) times higher Deprivation Index (NZDep2001), a small than for European children, while rates for area index of deprivation. The NZDep2001 Pacific children were 1.76 (95% CI 1.32–2.33) is constructed from nine socio–economic times higher. Admissions were also 1.24 variables from the 2001 census, covering (95% CI 1.04–1.48) times higher for males multiple factors.13 The index is traditionally than for females.
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 12 www.nzma.org.nz/journal ARTICLE
Table 1: Acute hospital admissions for otitis media in children in New Zealand for the time period of 2002 to 2008. Rate Ratio (RR) reference categories are NZDep Decile 1–2, European and Female: Note RR are unadjusted. Age 0–4 years Age 5–14 years rate per 1,000 RR 95% CI rate per 1,000 RR 95% CI New Zealand Deprivation Index Decile 1-2 1.07 1.00 0.07 1.00 3-4 1.46 1.36 1.19–1.55 0.08 1.16 0.81–1.66 5-6 1.55 1.44 1.27–1.64 0.10 1.49 1.06–2.11 7-8 1.96 1.83 1.61–2.06 0.13 1.87 1.35–2.58 9-10 2.80 2.61 2.32–2.92 0.20 2.93 2.17–3.95 Prioritised Ethnicity European 1.59 1.00 0.09 1.00 Māori 2.36 1.49 1.38–1.60 0.18 1.95 1.60–2.37 Pacific 2.43 1.53 1.38–1.69 0.16 1.76 1.32–2.33 Asian/ 1.05 0.66 0.56–0.77 0.07 0.72 0.47–1.48 Indian Gender Female 1.65 1.00 0.11 1.00 Male 2.06 1.25 1.17–1.33 0.13 1.24 0.47–1.10
Table 2: Elective admissions for ventilation tubes in children in New Zealand for the time period of 2002–2008. Rate Ratio (RR) reference categories are NZDep Decile 1–2, European and Female: Note RR are unadjusted.
Age 0-4 years Age 5-14 years rate per 1,000 RR 95% CI rate per 1,000 RR 95% CI New Zealand Deprivation Index Decile 1-2 8.61 1.00 1.97 1.00 3-4 10.28 1.19 1.14–1.25 2.77 1.40 1.31–1.50 5-6 12.47 1.45 1.38–1.52 3.90 1.98 1.86–2.10 7-8 13.62 1.58 1.51–1.65 5.08 2.57 2.43–2.72 9-10 12.59 1.46 1.40–1.53 6.38 3.23 3.06–3.41 Prioritised Ethnicity European 13.04 1.00 3.27 1.00 Māori 11.58 0.89 0.86–0.92 5.69 1.74 1.68–1.80 Pacific 9.71 0.74 0.71–0.78 7.33 2.24 2.15–2.34 Asian/ 3.42 0.26 0.24–0.29 1.32 0.40 0.37–0.44 Indian Gender Female 9.57 1.00 3.63 1.00 Male 13.76 1.44 1.40–1.48 4.62 1.27 1.23–1.31
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 13 www.nzma.org.nz/journal ARTICLE
There were also significant social gradients rates were 1.74 (95% CI 1.68–1.80) times in acute hospital admissions for otitis media higher, while for Pacific children rates were in both age groups. For example children 2.24 (95% CI 2.15–2.34) times higher. Admis- aged 0–4 years living in the most deprived sions were also 1.27 (95% CI 1.23–1.31) areas (NZDep2001 deciles 9–10) had 2.61 times higher for males than for females. times (95% CI 2.32–2.92) higher admission During 2002–2008, a social gradient in rates than those living in the least deprived elective admissions for ventilation tubes (NZDep2001 deciles 1–2) areas. was observed in both age groups. For Table 2 presents elective admissions for example, children aged 0–4 years living in the insertion of ventilation tubes during the most deprived areas (NZDep2001 decile the same time period. In the 0–4-year-old 9–10) were 1.46 (95% CI 1.40–1.53) times age group, Māori and Pacific children more likely to be admitted for ventilation had significantly lower ventilation tube tubes than those in the least deprived areas admission rates than did European (NZDep2001 decile 1–2). children. The rate ratio was 0.89 (95% CI Figure 1 summarises the relationship 0.86–0.92) for Māori children and 0.74 (95% between age and ethnicity for both acute CI 0.71–0.78) for Pacific children. Admission hospital admissions for otitis media and rates were 1.44 (CI 1.40–1.48) times higher elective admissions for ventilation tubes. In for males than for females. this figure, acute admissions for otitis media In contrast, in the 5–14 year old age were higher for Māori and Pacific children group, Māori and Pacific children had than for European children during the first significantly higher rates of admission 4 years of life, with the highest rates being for ventilation tube insertion than did seen in children at one year of age. Small European children. For Māori children, numbers after 3 years of age made interpre-
Figure 1: Acute Hospital Admissions for Otitis Media vs. Elective Admissions for Ventilation tubes in Children 0–14 Years by Age and Ethnicity, New Zealand 2002–2008
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 14 www.nzma.org.nz/journal ARTICLE
Figure 2: Elective Admissions for Ventilation tubes in 0–14 Year Olds by Age Group, Ethnicity and New Zealand Deprivation Index, New Zealand 2002–2008.
tation of ethnic differences by single year of 2007–2008. During this period, Māori and age less reliable for older children. European children aged 0–4 years had In contrast elective admissions for venti- similar first ENT appointment rates, with lation tubes were higher for European rates for Pacific and Asian/Indian children children than for Māori, Pacific and Asian being much lower. In contrast, Māori children during the first 5 years of life. At 5 children aged 5–14 years had the highest years and older the pattern reversed, with first appointment rates, followed by Māori and Pacific children having higher Pacific, European and then Asian/Indian admission rates than European children. children. The 0–4 year age group also had a higher total first appointment rate than Figure 2 shows elective admissions for the 5–14 year age group. ventilation tubes by age group, NZ Depri- vation Index quintile and ethnicity. For those aged 0–4 years, social gradients were evident for European and Māori children, Table 3: Attendance of children at first Ear with admissions increasing with increasing Nose and Throat Outpatient Clinics by age, NZDep2001 deprivation. For Pacific and New Zealand during 2007–2008. Asian children however, no such social Ethnicity Attended Did not Attend or gradients were evident. Further, in the more (%) Did not Wait (%) deprived areas (NZDep2001 decile 7–10) Aged 0–4 years admissions were much higher for European, European 92.5 7.5 then Māori, then Pacific, then Asian/Indian children. For children aged 5–14 years Māori 77.8 22.2 however, social gradients were evident for Pacific 84.2 15.8 all ethnic groups, with admissions increasing Asian/Indian 92.0 8.0 with increasing NZDep2001 deprivation. Aged 5–14 years At each level of deprivation, admissions European 91.7 8.3 were higher for Pacific, then Māori, then European, then Asian/Indian children. Māori 79.2 20.8 Figure 3 explores first ENT outpatient Pacific 84.0 16.0 appointment rates by ethnicity during Asian/Indian 93.1 6.9
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 15 www.nzma.org.nz/journal ARTICLE
Figure 3: Rate of referral for first appointment to Ear Nose and Throat Outpatient Clinics by age and ethnicity in New Zealand children during 2007–2008.
Table 3 reviews the attendance and Pacific children in their samples.16 The non-attendance rates of children at their ethnic differences seen in our study are in first ENT outpatient appointment during keeping with the results of a study by Milne 2007–2008. During this period, Māori and Vander Hoorn,17 as well as with local children had the highest non-attendance hearing screening data.5 This study found for their first appointment (22.2% at 0–4 higher tympanometry failure rates at 3 years and 20.8% at 5–14 years) followed by years and audiology and/or tympanometry Pacific children (15.8% and 16.0% respec- failure rates at 5 years for Pacific and tively). European and Asian children had Māori children than for European/ Pākehā the lowest non-attendance rates. children.5 While failed tympanometry and audiology are only indirect markers of OM Discussion prevalence, they do nevertheless suggest Part one of this study examined acute a higher prevalence of middle ear disease hospital admissions for otitis media and in Māori and Pacific children. Similarly a their distribution by gender, ethnicity study by Paterson et al, found a high popu- and socio-economic status. It found that lation prevalence of OME in Pacific Island 18 admissions occurred most commonly in 2-year-olds (25.4%). Higher rates of OM 1-year-olds, followed by infants, with rates are also present in other indigenous popula- being lowest in school-age children of all tions such as Australian Aboriginal children, 19,20 ethnic groups. The predominance of admis- American Indian and Inuit children. sions in the younger age groups is similar to By contrast, in our study Asian/Indian international studies, which show that the children were the least likely group to be peak incidence of AOM is approximately 6 admitted acutely with otitis media. It is months to 1 year of age.1,10,14 unclear whether this reflects a lower prev- In this study, acute admissions for alence of OM or a lower rate of hospital otitis media also varied by ethnicity, with presentation (vs. attending the GP). 21 Māori and Pacific children having higher However, one prospective US cohort study admission rates than European children in examining risk factors for OM also found both age groups. While hospitalisation data that after adjusting for confounders, Asian is a poor proxy for assessing ethnic differ- infants were less likely to be diagnosed with ences in AOM incidence, few other reliable OM than White Infants (OR 0.77 95% CI data sources are available. Those local 0.57-1.00). In this study, the reason for this studies which have explored the incidence disparity was unknown. of AOM either do not report ethnic specific Research suggests that the prevalence rates,15 or have low numbers of Māori and of OME follows that of AOM, with a peak
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 16 www.nzma.org.nz/journal ARTICLE
at 2 years of age and a smaller peak at 5 the results do suggest is that European years.21,22 Our analysis by single year of children living in average to more age showed that in European, Māori and deprived areas (NZDep2001 decile 5–10) Asian children, ventilation tubes were most are accessing ventilation tubes at much commonly inserted in 1-year-olds, with a higher rates than Māori or Pacific children, second smaller peak at 4 years of age. This with the lowest access rates being seen in is similar to the ages of peak prevalence for Asian children. Given that a much higher OME. For Pacific children however, the peak proportion of Māori and Pacific children occurred at 6 years, with a second smaller live in the most deprived NZDep areas,27 peak at 1 year of age. While the reasons for the magnitude of these inequalities in this are unclear, it is possible that the peak access is particularly concerning. at 1 year1,2,10 follows the peak incidence in In contrast, for children aged 5–14 years, AOM while the peak at six years reflects social gradients were evident for all ethnic delays in access to hearing screening. For groups. Further, at each level of NZDep2001 example, hearing screening (tympanometry) deprivation, admissions were higher for at 3 years of age during the study period was Pacific, then Māori, then European, then often undertaken in preschool settings, with Asian/Indian children. The finding that Pacific children having known lower rates access to ventilation tubes better matched of prior engagement with early childhood need in this older age group potentially education at school entry.23 suggest that routine Well Child hearing When aggregated into broader age cate- screening may be playing a role in identi- gories, Māori and Pacific children aged 0–4 fying unmet need in older children. years were significantly less likely to be admitted electively for ventilation tubes While hospital admission data is useful than European children, despite having for identifying inequalities in ventilation a likely greater burden of middle ear tube access, it is unable to shed any light disease. The pattern was reversed for 5–14 on the pathways leading to the paradoxical year olds, with Māori and Pacific children access patterns seen in younger children. In having significantly higher ventilation tube particular, it is unclear whether Māori and admission rates than European children. Pacific children had lower ventilation tube access rates because: they were referred less Lower rates of ventilation tube insertion often from primary care to otolaryngologists; in indigenous children and those from once referred they attended outpatient clinics ethnic minority groups have been reported less often; or once assessed, otolaryngologists in other countries.24,25,26 In Western admitted Māori and Pacific children less Australia, Spilsbury found that in the 0–4 year old age group, the lowest rates of often for ventilation tube insertion. Further, ventilation tube insertion were in indig- the impact the underreporting of privately enous girls and the highest rates were in funded, private hospital admissions to the non–indigenous boys. They also found that NMDS has on the findings is unclear (as many the age of first ventilation tube insertion children admitted privately for ventilation for indigenous children was on average tube insertion do not appear in the NMDS). 20 months older than for non–indigenous As the outpatient data in this study was not children. Spilsbury concluded this indicated diagnostically coded (only a health specialty reduced access to health care especially code is available), our analysis of first ENT given the high rate of OM seen in the appointments was only able to provide partial Australian indigenous population.24 answers to the questions raised above. This In our study, access to ventilation tubes is because many children attend ENT clinics also varied with neighbourhood depri- for other reasons (eg assessment for tonsil- vation, with admission rates for Māori lectomy), with the authors being unable to and European children aged 0–4 years filter their data from the analysis. However, increasing with increasing NZDep2001 given that ventilation tube insertion is the deprivation. For Pacific and Asian/Indian leading reason for elective ENT admissions in children however, no such social gradients New Zealand children,28 the contribution such were evident. While the reasons for these referrals make to ENT first appointments is differential gradients are unclear, what likely to be significant.
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 17 www.nzma.org.nz/journal ARTICLE
Even given these limitations, the finding that Māori and Pacific children aged Conclusions 0–4 years had similar or lower first ENT This study highlights large ethnic differ- appointment rates than European children, ences in access to ventilation tubes amongst despite having a likely higher burden of New Zealand’s 0–4 year olds, with the middle ear disease is concerning. Such greatest disparities being seen for Māori, findings potentially suggest that barriers in Pacific and Asian/Indian children living access may be occurring at the primary care in the most deprived areas. For Māori and level. Further, once referred, the higher Pacific children, such differences cannot be non-attendance rate of Māori and Pacific attributed to lower rates of AOM or OME children at their first appointments also when compared to European children. raises questions about barriers to access (eg The fact that similar patterns are not seen travel and parking costs, time off work) to for older children (5–14 years) potentially outpatient clinics. suggests that routine Well Child hearing Such concerns are supported by the 2004 screening may be playing a role in identi- New Zealand Living Standards survey fying unmet need in this older age group. which found that Māori, when compared to The disparities seen also suggest that European, were more likely to report a time factors, over and above OM prevalence, in the past when they needed to access the may be influencing access to ventilation GP but did not do so. The main reason for tubes in younger children. In particular, this in the survey was appointment costs.29 further research is required to assess why Another study in South Auckland investi- Māori and Pacific children aged 0–4 years gating non-attendance at outpatient clinics have similar or lower first appointment also highlighted cost, alongside multiple rates at ENT outpatient clinics than factors such as information and transport.30 European children, despite having a higher This study found in 1998 the average non-at- burden of middle ear disease. Further tendance rate of Māori patients to outpatient research is also warranted into the reasons appointments at South Auckland Health was for higher non-attendance rates of Māori 23%, which is similar to the figure seen in this and Pacific children at ENT outpatient study. Paterson et al, in their study, also found clinics. While the specific nature of barriers similar high non-attendance rates for Pacific to access is unknown, issues such as cost, children.18 transport and parents needing to take time Given the limitations of the outpatient of work may play a role. datasets however, it is likely that more defin- Given the importance of early detection itive answers will need to await the diagnostic and treatment of OM for children’s ongoing coding of outpatient data, so that the referral well-being and education, a greater pathways from primary care, through ENT understanding of the reasons for these clinics, to surgery can better be assessed. inequalities is urgently required.
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 18 www.nzma.org.nz/journal ARTICLE
Competing interests: Dr McCallum reports while undertaking this dissertation she was training as a Public Health physician and was receiving a bursary from the NZ College of Public Health. She had non- financial support from the NZCYES during the conduct of the study. Dr Craig reports DHB and Ministry of Health Funding for NZ Child and Youth Epidemiology Service (NZCYES) Reporting: This paper utilises a similar methodology to that used by the NZCYES for its reporting of hospitalisations–but extends it with new material on outpatients data. Some of Dr Craig’s salary was funded from NZCYES contracts with the MoH and DHBs, however the MoH and DHBs had no input into the methodology used to report the data. Dr Craig was a political candidate for Labour at the 2014 election–the paper was written prior to this and the paper does not make any recommendations of a political nature. Acknowledgments: Andrew Wicken, NZ Child and Youth Epidemiology Service, Dunedin for checking all data. Author information: Justine S McCallum, Paediatric Registrar, Dunedin Hospital, Dunedin; Liz Craig, Senior Research Fellow, NZ Child and Youth Epidemiology Service, University of Otago, Dunedin; Ian Whittaker, Postdoctoral Research Fellow, Space Physics Group, Department of Physics, University of Otago, Dunedin; Jo Baxter, Associate Dean for Māori Health and Director of the Centre for Hauora Māori, Dunedin Corresponding author: Justine S McCallum, Paediatric Unit, Dunedin Hospital, Great King Street, Dunedin, Otago, 9016. [email protected] URL: www.nzma.org.nz/journal/read-the-journal/all-issues/2010-2019/2015/ vol-128-no-1416-12-june-2015-/6547
REFERENCES: 1. Teele DW, Klein JO Vision and Hearing Jackson C (2008) The and Rosner B (1989). Screening Report – July Health Status of Children Epidemiology of Otitis 2005 to June 2006. www. and Young People in Media during the 1st 7 grevilleconsulting.co.nz Waitemata DHB. New Zealand Child and Youth years of life in children 6. Well Child Tamariki Ora Epidemiology Service. in Greater Boston – A Handbook www.heal- prospective, cohort study. thed.govt.nz/ resource/ 10. Casselbrant ML, Mandel Journal of Infectious well-child-tamariki- EM, Kurs-Lasky M, Rockette Diseases, 160(1), 83–94. ora-my-health-book HE, and Bluestone CD (1995). Otitis media in 2. Rovers MM, Schilder AGM, 7. Cook L (1998). Screening a population of black Zielhaus GA and Rosenfeld programmes for the detec- American and white RM (2004). Otitis Media. tion of otitis media with American infants, 0-2 years Lancet, 363(9407), 465–473. effusion and conductive of age. Int. J. Ped. Otorhi- 3. Glasziou PP, Del Mar CB, hearing loss in pre–school nolaryngology. 33 1-16 Sanders SL and Hayem M and new entrant school 11. Paradise JL, Rockette HL, (2004). Antibodies for acute children – an appraisal of Colborn DK, Bernard BS, otitis media in children. the literature. New Zealand Smith CJ, Kurs-lasky M, Cochrane Database Health Technology Assess- and Janosky JE (1997). Syst. Rev. (CD000219) ment (NZHTA) Report 3. Otitis media in 2253 4. Lous J, Burton MJ, Felding 8. Ministry of Health, N.Z. Pittsburgh-area infants: MJ, et al. (2005). Grommets Universal newborn hearing Prevalence and risk (ventilation tubes) for screening for New Zealand factors during the first hearing loss associated with 2005 – A report of the two years of life. Pedi- otitis media with effusion universal newborn hearing atrics 99(3), 318-333. in children. Cochrane Data- screening advisory group to 12. Mahadevan M, Navar- base Syst. Rev. (CD000219) the National Screening Unit, ro-Locsin G, Tan H K, 5. Greville Consulting. (2006). www.moh.govt.nz Yamanaka N, Sonsuwan (2006). New Zealand 9. Craig E, Anderson P and N, Wang PC, Dung NT,
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 19 www.nzma.org.nz/journal ARTICLE
Restuti RD, Hashim SS, and in 2 yr old Pacific Children children. Archives of Vijayasekaran S (2012). A living in New Zealand. Int. Otolaryngology–Head review of the burden of J. Ped. Otorhinolaryngology. & Neck Surgery. Vol disease due to otitis media 70, 1771-1778. 132, iss:11 doi: 10.1001/ in the Asia-Pacific. Int J 19. Morris P S, Leach A J, archotol.132.11.1216. Pediatr ORL, 76(5); 623-635 Silberberg P, Mellon G, 25. Kogan MD, Overpeck MD, 13. Tobias M, and Yeh LC, Wilson C, Hamilton, et Hoffman HJ, Casselbrant (2007). How much does al, (2005). Otitis Media in ML (2000), Factors associat- health care contribute young Aboriginal children ed with tympanostomy tube to health inequality in from remote communities insertion among pre-school New Zealand? Aust NZ in Northern and Central aged children in the J Public Health, 2007, Australia: a cross-sectional United States. Am. J. Public Jun;31(3):207–10. survey. BMC Pediatr, 5, 27. Health, 90(2), 245-250. 14. Homøe P, Christensen R 20. Curns A T, Holman R 26. Falster K, Randall D, Banks B and Bretlau P (1999). C, Shay D K, Cheek J E, E, Eades S, Gunasekera H, Acute otitis media and Kaufman S F, Singleton R J, Reath J, and Jorm L (2013). age at onset among et al. (2002). Outpatient and Inequalities in ventilation children in Greenland. Acta hospital visits associated tube insertion procedures Otolaryngol. 119(1) 65-71 with otitis media among between Aboriginal and American Indian and 15. Gribben B, Salkeld L J, non-Aboriginal children Alaska native children Hoare S, Jones H F (2012). in New South Wales, younger than 5 years. The incidence of acute Australia: a data linkage Pediatrics, 109(3), E41-41. otitis media in New study. BMJ Open 2013 Zealand under five years 21. Vernacchio L, Lesko SM, 3(e00307) doi:10.1136/ of age in the primary care Verzina RM, Corwin MJ, bmjopen-2013-003807 setting. J. Prim. Health Hunt CE, Hoffman HJ, 27. Robson B and Harris R Care, 4(3) 205-212. Mitchell AA. (2004) Racial/ (2007) Hauora: Māori ethnic disparities in the 16. Chalmers D (1989). Otitis Standards of Health diagnosis of OM in infancy. Media with Effusion in IV – A study of the years International Journal of children, the Dunedin 2000–2005. Wellington, study. Volume 108 of Pediatric Otorhinolaryn- Ministry of Health. clinics in developmental gology, 68(6): 795- 804. medicine. University of 22. Midgely EJ, Dewey C, Pryce 28. Craig E, Jackson C, Han D, Michigan, Mac Keith Press. K and Maw AR (2000). et al. Monitoring the health of New Zealand children 17. Milne RJ, and Vander The frequency of otitis and young people. Auck- Hoorn S (2010) Burden and media with effusion in land: Paediatric Society Cost of Hospital Admissions British pre–school children: of New Zealand, New for Vaccine–Preventable A guide for treatment. Clini- Zealand Child and Youth Paediatric Pneumococcal cal Otolaryngology, vol 25, Epidemiology Service, 2007 Disease and Non–Typable iss:6. Doi: 10.1046/j.1365– Haemophilus influenza 2273.2000.00360.x 29. Ministry of Social Develop- Otitis Media in New 23. New Zealand Childcare ment (2004). New Zealand Zealand. Appl. Health Econ Survey 2009 (Revised 17 living standards 2004. Health Policy, 8(5): 281–300 December 2010) Highlights, http://www.msd.govt.nz 18. Paterson J E, Carter S, Statistics New Zealand 30. Coffin TK, Emery B, Zwier Wallace J, Ahmad Z, Garrett 24. Spilsbury K, Kadhim A L, G (2000) Report on reasons N, and Silva P A (2006). Semmens J B and Lannigan for Māori non–attendance Pacific Island families F J (2006). Decreasing rates of outpatient clinic appoint- Study: the prevalence of of middle ear surgery ments. Counties Manukau chronic middle ear disease in Western Australian District Health Board.
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 20 www.nzma.org.nz/journal ARTICLE
Whooping cough—where are we now? A review Tomasz Kiedrzynski, Ange Bissielo, Mishra Suryaprakash, Don Bandaranayake
ABSTRACT AIMS: This paper describes the recent trends of pertussis and vaccine uptake in New Zealand based on notifications and immunisation registration information since 2011. It highlights the current risk for the infant in the first months after birth and the crucial role a pertussis booster in pregnancy could play. It also aims to show that protection of infants by the acellular pertussis vaccine can be improved by timely immunisation even in a situation of improving overall uptake rates that are nearing the national target of 95%. METHODS: We analysed New Zealand notification data for pertussis, extracted from EpiSurv between August 2011 and December 2013, which included the period of the last epidemic. Pertussis immunisation coverage data were extracted from the National Immunisation Register (NIR). Population estimates were based on 2006 census data. Deprivation was analysed using the New Zealand Deprivation Index 2006. RESULTS: Despite immunisation coverage at 12 months having exceeded 90% New Zealand experienced a large epidemic from 2011 to 2014, with several hundred infant hospitalisations and three deaths. Notification data indicated an average annual rate of pertussis in the New Zealand population of 102 per 100,000 with the highest rates in the youngest age groups. While an overall increase in immunisation coverage in New Zealand was evident and the timeliness showed improvement across ethnic groups and deprivation deciles, there was a marked geographical variation within DHBs and between ethnic groups. CONCLUSIONS: Given the recent published evidence, pertussis vaccination should be offered to all mothers between weeks 28 and 38 of pregnancy. Further improvements are still possible in coverage at 6 months, particularly in Māori and but also in Pacific populations, as well as in more deprived populations. DHBs work towards achieving the 95% target can contribute to the improvement in the timeliness of immunisation.
and less effective at preventing infection Introduction with and transmission of Bordetella Whooping cough (pertussis) is a disease pertussis.2 affecting all age groups and is often severe The main goal of the pertussis immu- in young children, most notably in infants. nisation programme is to protect infants Pertussis is both an endemic and epidemic disease, with 2-to-5 yearly epidemic cycles. in the first year of life against severe, Epidemics last up to 3 years (Figure 1). life-threatening disease. While timely immunisation is crucial, the first dose of Acellular pertussis vaccines replaced the vaccine is not given until 6 weeks of the whole cell vaccine in the New Zealand age, leaving a window of highest risk to schedule in 2000. In comparison with children who have received an immuni- the young infant. The risk of infection sation schedule that includes both whole in infants has however been shown to cell and acellular vaccine, children who decrease significantly after the first and the have received only acellular pertussis second doses of the vaccine.3,4 The World vaccine appear at increased risk of Health Organization (WHO) considers that pertussis.1 Thus, the acellular pertussis in general, at least two doses of acellular vaccines currently in use appear to be less vaccine are required for protection, and immunogenic than the whole cell vaccines recommends a three-dose primary series,
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 21 www.nzma.org.nz/journal ARTICLE
Figure 1. Number of pertussis notifications and hospitalisations by calendar month-year, January 1998 to September 2014
Note: Includes confirmed, probable, suspect cases and notifications that were still under investigation. Source: ESR, EpiSurv which is given at 6 weeks, 3 months and 5 from August 2011 to December 2013, which months in New Zealand.5,6 was the peak activity of the 2011–2014 To address this immunity gap between pertussis epidemic. We included confirmed, birth and the acquisition of vaccine-in- probable, and suspect cases as defined in duced immunity, maternal pertussis the Ministry of Health’s Communicable immunisation between weeks 28 and 38 Disease Control Manual 2012. Hospital- of pregnancy is now recommended.7 In isation information and immunisation New Zealand this intervention has been status of cases at any time prior to onset funded since 1 January 2013. The expected of disease were as recorded in Episurv effect is protection of the infant’s mother (Health professionals use a range of sources against pertussis and passive transfer of to update immunisation status including protective amounts of maternal antibodies the NIR, parental recall or Well Child book to the infant. A recent study has shown a records. This does not take into account vaccine efficacy of 91% for infants younger the time needed for the vaccine to elicit an than 3 months following vaccination during immune response). pregnancy.8 The safety of the pertussis and Pertussis immunisation coverage data tetanus toxoid, reduced diphtheria toxoid, were extracted from the National Immu- and acellular pertussis (Tdap) vaccine in nisation Register (NIR) on 12 August 2014. pregnancy has also been demonstrated.9,10 They included the numbers of eligible An additional strategy to prevent whooping children completing 6, 8, 12 and 24 months cough in the new-born, ‘cocooning’, and their immunisation status for pertussis involves immunising those living with or at these ages. Coverage was the proportion caring for infants—particularly the parents of eligible children who were immunised. and other household contacts who are most Current District Health Board (DHB) of often the source of infant infections. These domicile was based on the National Health vaccinations are currently recommended Index. Progress in the timeliness of immuni- but not funded in New Zealand as the sation against pertussis was measured using degree of their effectiveness is unclear.11 the coverage by three doses of pertussis vaccine of children at 6 and 12 months Methods in the 3-month reporting periods ending We analysed New Zealand notification March 2011 and December 2013. data for pertussis extracted from EpiSurv Population estimates were based on 2006 as of 17 April 2014. Episurv is the national population census data. Ethnicity was prior- database for notifiable disease surveillance. itised. Deprivation was analysed using the The period of interest for this review was NZ Deprivation Index 2006 (NZDep2006).
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 22 www.nzma.org.nz/journal ARTICLE
Table 1. Pertussis vaccine three dose coverage at 6 and 12 months and pertussis incidence rates by deprivation quintile, August 2011–December 2013
Coverage at Annual incidence in infants Coverage 12 NZDep2006 6 months under 1 year of age per 100,000 months NZDep 1–2 81% 278 93% NZDep 3–4 80% 407 93% NZDep 5–6 78% 480 93% NZDep 7–8 75% 578 92% NZDep 9–10 67% 788 90%
Sources: ESR, EpiSurv and Ministry of Health, National Immunisation Register
The population aged less than 1 year was calculated using NZDep2006-specific popu- Deprivation and lation proportions in the 0-4 population timeliness of applied to the known total population in that age group. 3-dose pertussis We performed descriptive analyses to immunisation derive rates, case frequencies and immuni- of infants sation coverage. Table 1 shows that coverage with the three dose pertussis vaccine in the Results period August 2011 to December 2013, Notification data from peak activity of the particularly at 6 months, decreases while 2011–2014 pertussis epidemic indicated an pertussis incidence in infants increases average annual rate of notified pertussis with increasing household deprivation. in the New Zealand population of 102 per Between the 3 month reporting periods 100,000. A total of 778 cases were reported in ending in March 2011 and December 2013 infants under 1 year and 386 cases in the age the three dose pertussis immunisation group 12–23 months. The average annual coverage at the age of 6 months, while incidence rates were highest in the youngest remaining low in deprivation levels 7–8 age groups: 801 per 100,000 under 6 months, and 9–10, has improved from 68 and 60% 396 in 6 to 7 months, 309 in the 8 to 11 to 77 and 69% respectively. The coverage at months and 293 in the 12 to 23 months age 12 months was higher, and increased from 89 and 87% respectively, to 94% for both groups. Three deaths were also notified (2 deprivation levels. infants under 1 year and a 3-year-old child). If we look more specifically at the popula- tions that reached 6, 8, 12 and 24 months of Ethnicity and age during the period from August 2011 to timeliness of December 2013, immunisation coverage for 3-dose pertussis three doses of pertussis vaccine increased from 75% at the age of 6 months to 88% at immunisation 8 months, 92% at 12 months and 95% at 24 of infants months. This shows that not enough infants Table 2 highlights that ethnicities with were being immunised early enough during higher ‘on-time’ coverage for the three this period. doses of pertussis vaccine had a lower However, the timeliness of the three-dose pertussis incidence in infants: Asian infants immunisation against pertussis has had the highest coverage at 6 months and improved during the period between 2011 the lowest incidence rate. At the opposite and 2013. The coverage of children with end of the range, Māori and Pacific infants three doses of the vaccine at the age of 6 had the lowest coverage and the highest months has gradually increased from 67 to incidence. While Pacific infants had better 78% between the 3-month reporting periods coverage than Māori infants at 6 and 12 ending March 2011 and December 2013. months, their pertussis incidence was the Similarly, this coverage at age 12 months highest likely due to other factors, eg envi- has gradually increased from 89 to 94%. ronmental, such as overcrowding.
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 23 www.nzma.org.nz/journal ARTICLE
Table 2. Pertussis vaccine 3-dose coverage at 6 and 12 months and pertussis incidence rates by prioritised ethnicity, August 2011–December 2013 Ethnicity Coverage at Annual incidence in infants under Coverage at (prioritised) 6 months 1 year of age per 100,000 12 months Asian 88% 226 96% NZE 81% 520 93% Other 76% 485 88% Pacific 73% 934 94% Māori 62% 734 89%
Sources: ESR, EpiSurv and Ministry of Health, National Immunisation Register
Figure 2. Immunisation coverage for 3 doses of pertussis vaccine at ages 6 and 12 months by DHB, August 2011 to December 2013, New Zealand
Source: Ministry of Health, National Immunisation Register
Between the 3 month reporting periods Immunisation coverage at age 12 ending in March 2011 and December 2013 months was higher and ranged from 86% the three-dose pertussis immunisation in Northland to 95% in Hutt Valley, with coverage at the age of 6 months in Māori and Northland, West Coast, Waikato and Bay of Pacific infants had improved from 52 and Plenty being the only DHBs with coverage at 66% to 64 and 76% respectively. Similarly, age 12 months that was <90%. coverage at 12 months had increased from 83 and 92% respectively, to 92 and 97%. Pertussis incidence Geography and and immunisation timeliness of status in children under 2 years of age 3-dose pertussis Figure 3 shows the immunisation status of immunisation children with notified pertussis and Figure 4 shows the immunisation status of children of infants hospitalised with pertussis by age for the During the period from August 2011 to period 1 August 2011 to 31 December 2013. December 2013, the three-dose pertussis After a peak at 1 month of age, the number immunisation coverage at 6 months of of cases and hospital admission decreases age within DHBs varied between 60% in older age groups. Notified cases aged in Northland to 84% in Southern. DHBs 1 month or less were too young to be with the lowest coverage (under 70%) immunised. Fifty-six percent of cases aged were Northland, Waikato, Bay of Plenty, 6 months at the time of notification had Tairawhiti, and Whanganui (Figure 2). received three doses of vaccine.
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 24 www.nzma.org.nz/journal ARTICLE
Figure 3. Immunisation status of children with pertussis by age-in-months, during August 2011 to December 2013, New Zealand.
Source: ESR, EpiSurv
Figure 4. Immunisation status of children hospitalised with pertussis by age-in-months, during August 2011 to December 2013, New Zealand.
Source: ESR, EpiSurv
Figure 3 and Figure 4 show that the bulk Currently the uptake of pertussis vaccine of cases and hospital admissions are in during pregnancy in New Zealand is infants too young to be fully immunised poor, in spite of the recommendations with the primary three-dose immuni- from the CDC7 and strong evidence of its sation series. However, it is likely that a efficacy and safety.8 DHBs, with support substantial number of cases and hospital from the Ministry of Health, could take admissions in children aged over 5 months steps to improve the situation. This will could have been prevented if all vaccine undoubtedly help reduce disease incidence doses had been given on time.12 and severity in infants too young to be immunised. Conclusions Further improvements are still possible in There has been a substantial coverage at 6 months, particularly in Māori improvement in immunisation coverage and Pacific populations, as well as in popula- in New Zealand since 2011. While the tions living in more deprived households. protection offered by the acellular vaccine In July 2012 the government’s health target is not optimal, we assume a modified or was refocused on a 95% coverage at the new vaccine will not be available within the 8-month milestone age. Between July 2012 next 5 years. and December 2013, the 8 month coverage
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 25 www.nzma.org.nz/journal ARTICLE
progressed from 87 to 91% overall, from 78 to surge of pertussis should be—first of all— 88% in Māori, from 87 to 93% in Pacific, from pertussis vaccination of all, or most, mothers 87 to 91% in deprivation deciles 7–8 and from between weeks 28 and 38 of pregnancy, 81 to 90% in deprivation deciles 9–10. Further as well as immunisation coverage at and improvement in the timeliness of immuni- above 95% for the three doses of pertussis- sation is expected as DHBs work towards containing vaccine at age 6 months. achieving the 95% target. DHBs that have In 2015, the Ministry will be leading a difficulties in maintaining or improving their pertussis policy workshop with PHARMAC immunisation coverage will require addi- and The Immunisation Advisory Centre tional support from the Ministry of Health to support to build on existing programme help them to reach the target. strengths and review internal strategies for The priorities before the next epidemic reducing whooping cough.
Competing interests: Nil Acknowledgements: The authors thank Associate Professor Cameron Grant for comments on an earlier draft of this article Author information: Tomasz Kiedrzynski, Principal Advisor, Communicable Disease Team, Public Health, Clinical Leadership, Protection & Regulation, Ministry of Health, Wellington; Ange Bissielo, Senior Analyst (Epidemiology), Health Intelligence Team, Health Programme, Institute of Environmental Science and Research Ltd, Porirua; Mishra Suryaprakash, Senior Advisor (Epidemiology), National Immunisation Programme, Sector Capability and Implementation, Ministry of Health, Wellington; Don Bandaranayake, Public Health Physician, Health Intelligence Team, Health Programme, Institute of Environmental Science and Research Ltd, Porirua. Corresponding author: Don Bandaranayake, Institute of Environmental Science and Research Ltd, National Centre for Biosecurity and Infectious Disease, PO Box 40-158, Upper Hutt 5018 [email protected] Ph: (+61 4) 529 0627 Fax (+61 4) 978 6690 URL: http://www.nzma.org.nz/journal/read-the-journal/all-issues/2010-2019/ 2015/vol-128-no-1416/6559
REFERENCES: 1. Witt MA, Arias L, Katz 3. Juretzko P, Von Kries R, analyses of 10 years of PH, Truong ET, Witt DJ. Hermann M, von König pertussis surveillance. Reduced risk of pertussis CW, Weil J, Giani G. Vaccine. 2012; 30:3239-47. among persons ever Effectiveness of acellular 5. Bisgard KM, Rhodes P, vaccinated with whole cell pertussis vaccine assessed Connelly BL, et al. Pertussis pertussis vaccine compared by hospital-based active vaccine effectiveness to recipients of acellular surveillance in Germa- among children 6 to 59 pertussis vaccines in a large ny. Clinical infectious months of age in the United US cohort. Clinical infec- diseases. 2002; 35:162-7. States, 1998–2001. Pediat- tious diseases. 2013:cit046. 4. Nilsson L, Lepp T, von rics. 2005; 116:e285-e94. 2. WHO. Revised guidance Segebaden K, Hallander 6. WHO. Pertussis vaccines: on the choice of pertussis H, Gustafsson L. Pertussis WHO position paper. vaccines: July 2014. Weekly vaccination in infancy Weekly epidemiolog- epidemiological record. lowers the incidence of ical record. Geneva, Geneva, Switzerland: pertussis disease and the Switzerland: World World Health Organi- rate of hospitalisation Health Organization, zation, 2014; 337-44. after one and two doses: 2010; 385-400.
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 26 www.nzma.org.nz/journal ARTICLE
7. CDC. Updated Recom- observational study. The obstetrics and gynecology. mendations for Use of Lancet. 2014; 384:1521-8. 2012; 207:59. e1-. e7. Tetanus Toxoid, Reduced 9. Donegan K, King B, Bryan P. 11. Levy-Bruhl D. Protecting Diphtheria Toxoid, and Safety of pertussis vacci- the very young against Acellular Pertussis Vaccine nation in pregnant women pertussis–cough, costs and (Tdap) in Pregnant Women in UK: observational study. cocooning. Euro Surveill. — Advisory Commit- BMJ. 2014; 349:g4219. 2014; 19:pii. 20689. tee on Immunization 10. Zheteyeva YA, Moro PL, Practices (ACIP), 2012. 12. Grant CC, Roberts M, Tepper NK, et al. Adverse Scragg R, et al. Delayed MMWR, 2013; 131-5. event reports after immunisation and risk 8. Amirthalingam G, tetanus toxoid, reduced of pertussis in infants: Andrews N, Campbell diphtheria toxoid, and unmatched case-control H, et al. Effectiveness acellular pertussis vaccines of maternal pertussis in pregnant women. study. Bmj. 2003; 326:852. vaccination in England: an American journal of
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 27 www.nzma.org.nz/journal ARTICLE
The role of echocardiography in Staphylococcus aureus bacteraemia at Auckland City Hospital Nicholas Gow, Boris S. Lowe, Joshua Freeman, Sally Roberts
ABSTRACT AIMS: Current guidelines recommend echocardiography in all episodes of Staphylococcus aureus bacter- aemia (SAB). This study aimed to determine whether a very low-risk group of patients with SAB could be found in whom echocardiography was of no incremental diagnostic value for the diagnosis of infective endocarditis. METHODS: Using the ANZCOSS dataset, we identified 574 eligible episodes of adult SAB at Auckland District Health Board (ADHB) between 2007 and 2012, and retrospectively obtained additional microbio- logical and clinical data. Prevalence of IE was determined using the modified Duke’s criteria for diagnosis of IE. Multivariate logistic regression analysis was used to determine whether risk factors were independently associated with IE, and we also assessed their negative predictive value (NPV). RESULTS: Transthoracic and/or transoesophageal echocardiography was performed in 370 (65%) episodes of SAB. The prevalence of clinically definite and clinically possible IE was 5.6% and 8.5%, respectively. Thirty day all-cause mortality was 11.7%. The factors with the highest NPV when absent in hospital-acquired SAB were non central venous line-associated bacteraemia (100%), persistent bacteraemia (96%), and presence of a prosthetic valve or cardiac rhythm management (CRM) device (95%). When none of these three criteria were present the NPV was 100% (99-100%). CONCLUSIONS: A group of very low risk patients was found in our study: central line-associated SAB without prosthetic valves / CRM devices and without persistent bacteraemia. These patients had no episodes of IE and echocardiography is of no incremental diagnostic benefit.
taphylococcus aureus sepsis is a most common cause of probable or definite leading cause of community-onset endocarditis in 336 patients in which S and hospital-acquired sepsis in New the organism was known; 27% required Zealand with an estimated incidence surgical management.5 1 of 14/100,000/year hospital discharges. International guidelines for the diagnosis Infective endocarditis (IE) complicates and management of SAB recommend echo- S. aureus bacteraemia (SAB) in 6–30% of cardiography, preferably transoesophageal episodes with a 30 day mortality of 20%.2,3 (TOE) in all cases.6 Australasian guidelines S. aureus is the most common cause also recommend echocardiography in the of infective endocarditis globally,4 and workup of all cases of SAB to rule out IE, determining whether an episode of SAB is recommending a transthoracic echocar- complicated by endocarditis has important diogram (TTE) in all cases, with a TOE in prognostic and therapeutic implications. high risk patients or in selected patients In New Zealand, S. aureus was the second after negative or equivocal TTE.7 However
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 28 www.nzma.org.nz/journal ARTICLE
in a defined subgroup of patients with SAB The mode of onset of SAB was classified the use of echocardiography may not add as follows. Hospital onset healthcare-asso- predictive value to its exclusion.8–11 Echocar- ciated (HO-HA) infection was defined as a diography is a valuable resource requiring positive blood culture collected more than technical skill and clinical expertise. It 48 hours after admission or within 48 hours is widely available and has negligible of discharge, in patients without signs of risk making it prone to overuse in some infection on admission. Community onset, settings.12 It has an important role in the healthcare-associated (CO-HA) infection diagnosis and management of IE. However, was defined as patients with hospital- it is important to consider the pre-test prob- isation, surgery, dialysis, or long-term ability of IE as the sensitivity for both TTE care facility residence in the 12 months and TOE is less than 100%.13 prior to presentation. Community onset, The aim of this study was to review the community-associated (CO-CA) infection was community onset bacteraemia not results of echocardiography in Auckland meeting the above definition. Intravascular City Hospital patients with SAB to determine catheter-associated infection was defined the prevalence of infective endocarditis; as bacteraemia in which no other source to identify whether the clinical markers was identified and either the catheter tip positively and negatively associated with or insertion site cultured S. aureus or there IE in previous literature retained signifi- was resolution of signs and symptoms once cance in our setting; and if possible to use the line was removed. Peripheral line-asso- these markers to develop criteria optimising ciated SAB and haemodialysis intravascular patient selection for echocardiography. access device-associated infections were analysed separately to those associated with Methods a central venous line. Persistent bacter- Sequential episodes of SAB at Auckland aemia was defined as more than 96 hours City Hospital from May 2007 to December between the first positive and first negative 2012 were identified from data collected blood culture. We followed a previously locally for the prospective cohort study developed11 classification of “documented” Australia and New Zealand Collaboration persistent bacteraemia if there was a on Outcomes in Staphylococcal Sepsis positive blood culture collected between 48 (ANZCOSS).3 The entry criterion was at least and 96 hours, or “possible” persistent bacte- one blood culture positive with S. aureus raemia if not. Cases separated by 100 days in patients with clinical manifestations or more or with a differing antibiogram consistent with staphylococcal infection. were counted as a new bacteraemia.14 Exclusion criteria included patients less The modified Duke criteria were used to than 15 years of age; patients not admitted determine the diagnosis of definite (patho- or not treated for their bacteraemia; and logically proven or clinically definite) or those who were discharged, transferred, clinically possible IE.15 or who died in less than 48 hours. Data Univariate analysis was performed using collected included patient demographics, rank sum test for age and two-tailed Fisher’s date of admission, discharge or death, exact test for categorical data using two healthcare exposure, antimicrobial suscep- way contingency tables. Logistic regression tibility, device-related infections, and 7 and analysis was performed using SAS Institute 30 day all-cause mortality. For the present software to model the effects of factors study, additional information was obtained found to be significantly associated with from electronic and written clinical records: the primary outcome (IE) and to explore presence of prosthetic valves or cardiac their independence. Sensitivity, specificity, rhythm management (CRM) devices, 100 positive and negative predictive value for IE day all-cause mortality and readmission in SAB were calculated for selected clinical data, duration of bacteraemia (time to first criteria as follows: sensitivity (IE in episodes negative blood culture), and echocardiog- of SAB with the selected criteria / all raphy findings. 100 day follow-up data was episodes of IE), specificity (no IE without the used as a reference standard for the possible selected criteria / all cases SAB without IE), development of IE after discharge.14 positive predictive value (IE in episodes of
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 29 www.nzma.org.nz/journal ARTICLE
Figure 1: Flow diagram of episodes of SAB
SAB with the selected criteria / all episodes racic echocardiography (TTE) performed with the selected criteria), and negative in a median 3 (range 0–27) days; 28 had a predictive value (no IE in patients without second TTE in a median 5 (2–28) days; and the selected criteria / no IE without the 93 had transoesophageal echocardiography selected criteria + IE without the selected (TOE) in a median 7 (0–28) days. Of the criteria). A p-value < 0.05 was considered to 370 investigated with echocardiography, indicate statistical significance. Institutional 277 (75%) received TTE alone, 19 (5%) TOE ethics approval was obtained for a minimal alone, and 73 (20%) both modalities. 25 risk study. patients had a vegetation found at echo- cardiography, 4 had a vegetation and an Results abscess, and one patient had a perforated cusp. 372 patients (65%) had a follow-up 636 episodes of SAB in patients aged 15 blood culture within 96 hours. or over were encountered; 62 episodes Table 1 shows the baseline character- were excluded due to admission for less istics of the cohort included for analysis than 48 hours (42), no treatment (16), and including the characteristics of those a recurrent episode within 100 days (4) with and without echocardiography. 186 (Figure 1). The 4 episodes of recurrent SAB episodes (32.4%) were hospital-onset included a case of implantable port (Port- healthcare-associated and 252 (43.9%) were A-Cath®) infection which was not removed; community-onset healthcare-associated, recurrent septic arthritis in a patient with with the remainder community onset multiple areas of broken skin; a patient community-associated. The group with with a peripheral line infection readmitted echocardiogram performed had a higher with a secondary pelvic collection; and a proportion with documented persistent tunnelled haemodialysis line infection with bacteraemia (p <0.001), a higher proportion delayed treatment. None of these 4 patients of community onset SAB, both CO-HA (p = had endocarditis. 0.04) and CO-CA (p =0.024), injecting drug The remaining 574 episodes occurred users (p < 0.001), and haemodialysis line-as- in 554 patients. 14 patients had 2 episodes sociated SAB (p < 0.001); there were fewer of SAB over the 68 month period and 3 HO-HA episodes in this group. 7, 30 and 100 patients had 3 episodes. Echocardiography day mortality was higher in the group who was performed within 28 days in 370 did not receive echocardiography. 8 patients episodes of SAB (65%). 351 had transtho- (1.4%) were lost to 100-day followup.
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 30 www.nzma.org.nz/journal ARTICLE
Table 1: Characteristics of patients with SAB, by echocardiogram status Patient Total Echo performed Echo not performed p-value characteristics (n=574) (n=370) (n=204) Median age (IQR) 57 (43–71) 55.5 (41–69) 60 (45.75–75) 0.08
Male 342 (59.6%) 212 (57%) 130 (64%) 0.16
MRSA 40 (7.0%) 31 (8.4%) 9 (4.4%) 0.07
Prosthetic valve 28 (4.9%) 23 (6.2%) 5 (2.5%) 0.07
CRM device 23 (4.0%) 17 (4.6%) 6 (2.9%) 0.19
Persistent bacteraemia 306 (53.3%) 198 (54.5%) 108 (52.9%) 0.93
Documented persistent bacteraemia 78 (13.6%) 73 (19.7%) 5 (2.5%) <0.001
HO-HA 186 (32.4%) 97 (26.2%) 89 (43.6%) <0.001
CO-HA 252 (43.9%) 174 (47.0%) 78 (38.2%) 0.04
CO-CA 136 (34.7%) 99 (26.8%) 37 (18.1%) 0.024
IVDU 15 (2.6%) 15 (4.1%) 0 0.002
CVL-associated 89 (15.5%) 60 (16.2%) 29 (14.2%) 0.55
Haemodialysis line-associated 45 (7.8%) 40 (10.8%) 5 (2.5%) <0.001
7 day all-cause mortality 22 (3.8%) 5 (1.4%) 17 (8.3%) <0.001
30 day all-cause mortality 67 (11.7%) 34 (9.2%) 33 (16.2%) 0.02
100 day all-cause mortality 93 (16.2%) 48 (13.0%) 45 (22.1%) 0.006
Definitions: CRM = cardiac rhythm management; HO-HA = hospital-onset healthcare-associated; CO-HA = community-onset healthcare-associated; CO-CA = community-onset community-associated; IVDU = intravenous drug user; CVL = central venous line
There were 33 cases (5.6%) of ‘definite IE’; 2) in this cohort: CO-CA bacteraemia, 10 pathologically proven and 23 clinically ‘documented’ persistent bacteraemia, the definite cases according to the modified presence of a prosthetic valve or CRM Duke criteria. The 23 clinically definite device, and injecting drug use. No episodes cases met 2 major criteria in 19 episodes, of IE were encountered in patients with and 1 major and 3 or more minor criteria in CVL-associated SAB. Removal of the CVL was the remaining 4 episodes. The addition of 16 documented in 68 cases (76%), and occurred ‘clinically possible’ cases gave a total prev- in a median length of 1 day after positive alence of IE of 49 (8.5%), with 38 left sided blood culture (range 0–19 days, interquartile and 11 right sided. 2 of the 33 clinically range 1–2 days). These findings were definite cases did not have TOE changes consistent when analysing the 49 cases of consistent with endocarditis. definite and possible IE and were confirmed to be independently associated with IE after The sensitivity of TTE for the detection logistic regression analysis was performed of IE was 64% (16 abnormal TTEs in 25 (Table 3). The sensitivity, specificity, and patients diagnosed with IE); excluding positive and negative predictive values of patients with prosthetic valves or CRM selected clinical criteria for cases of definite devices did not increase the sensitivity. In IE are shown in Table 4. The high negative 6 patients with negative TTEs performed at predictive value for IE of CVL-associated SAB median 3 days (range 0-13), and 3 indeter- was further strengthened when combined minate TTEs performed at median 3 days with the absence of persistent bacteraemia, (range 3-6), subsequent TOE findings were and the absence of prosthetic valves or consistent with IE. No additional cases cardiac rhythm management devices. developed IE in the 100 day follow-up 80 episodes of CVL-associated SAB with period. In the cases with definite IE the 30 complete follow-up data met no high risk day all-cause mortality was 21%. criteria. The 100 day all-cause mortality was The following risk factors were signifi- 12.5% and they were investigated with 50 cantly associated with definite IE (Table TTEs and 8 TOEs.
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 31 www.nzma.org.nz/journal ARTICLE
Table 2: Factors associated with definite IE Risk All patients Patients with- Patients with OR (95% CI) p-value factor n=574 out definite IE definite IE n=541 n=33 HO-HA 186 (32.4%) 183 (33.9%) 3 (9.1%) 0.2 (0.1–0.68) 0.002
CO-HA 252 (43.9%) 239 (44.2%) 13 (39.4%) 0.82 (0.38–1.77) 0.718
CO-CA 136 (23.7%) 119 (22.0%) 17 (51.5%) 3.8 (1.8–8.1) <0.001
Persistent bacteraemia 306 (53.3%) 283 (52.3%) 23 (69.7%) 2.1 (0.9–4.8) 0.07
‘Documented’ persistent bacteraemia 78 (13.6%) 65 (12.0%) 13 (39.4%) 4.8 (2.1–10.6) <0.001
Prosthetic valve / CRM device 50 (8.7%) 37 (6.8%) 13 (39.4%) 8.85 (3.8–20.5) <0.001
Peripheral IV line-associated 65 (11.3%) 62 (11.5%) 3 (9.1%) 0.77 (0.18–2.76) 1.00
HD catheter-associated 45 (7.8%) 43 (7.9%) 2 (6.1%) 0.75 (0.12–3.36) 1.00
CVL-associated 89 (15.5%) 89 (16.5%) 0 0 (0–0.5) <0.001
IVDU 15 (2.6%) 11 (2.0%) 4 (12.1%) 6.65 (1.7–24.6) 0.008
Definitions: CI = confidence interval; CRM = cardiac rhythm management; HO-HA = hospital-onset heathcare-associated; CO-HA = community-onset healthcare-associated; CO-CA = community-onset community-associated; IVDU = intravenous drug user; CVL = central venous line
Table 3: Multivariate analysis Risk All patients Patients without Patients with Log p-value factor n=574 definite IE definite IE odds n=541 n=33 Community onset (CO-CA and CO-HA) 358 (62.3%) 328 (60.6%) 30 (90.9%) 3.6 0.002
‘Documented’ prolonged bacteraemia 78 (13.6%) 65 (12.0%) 13 (39.4%) 2.9 0.004
Prosthetic valve / CRM device 50 (8.7%) 37 (6.8%) 13 (39.4%) 5.24 0.003
CVL-associated 89 (15.5%) 89 (16.5%) 0 0 0.001
IVDU 15 (2.6%) 11 (2.0%) 4 (12.1%) 3.1 0.002
Definitions: CO-HA = community-onset healthcare-associated; CO-CA = community-onset community-associated; CVL = central venous line; CRM = cardiac rhythm management; IVDU = intravenous drug user
Table 4: Sensitivity, specificity, positive and negative predictive values for definite IE, stratified by selected clinical criteria Sensitivity Specificity Positive PV Negative PV (95% CI’s) Community onset (CO-CA and CO-HA) 91% (75–98%) 39% (38–40%) 8% (7–9%) 99% (96–100%)
Non CVL-associated 100% (87–100%) 17% (16–17%) 7% (6–7%) 100% (95–100%)
“Documented” prolonged bacteraemia 40% (23–58%) 88% (85–91%) 17% (9–27%) 96% (94–98%)
Prosthetic valve / CRM device 40% (23–58%) 93% (91–95%) 26% (15–40%) 95% (93–97%)
No criteria present* 100% (89–100%) 91% (89–94%) 41% (30–50%) 100% (99–100%)
Definitions: CI = confidence interval; CO-HA = community-onset healthcare-associated; CO-CA = community-onset community-associated; CVL = central venous line * ie, an episode of HO-HA CVL-associated SAB which does not persist >96 hours in a patient without prosthetic cardiac valves or a CRM device.
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 32 www.nzma.org.nz/journal ARTICLE
Our observations support the findings Discussion of previous studies which suggest that This large retrospective evaluation of SAB clinical factors may be used to determine at a tertiary referral centre in New Zealand a population of patients with SAB, partic- established a 5.6% prevalence of definite ularly in HO-HA infection, in which IE is IE, and 8.5% of definite and possible IE. No rare and the use of echocardiography can episodes of IE were detected in patients be rationalised. A series of 213 episodes with CVL-associated SAB. The negative of SAB8 found that the 3 cases of IE in 87 predictive value for episodes of HO-HA cases of catheter-related SAB had other CVL-associated infection without persistent recognised risk factors for IE; persistent bacteraemia and without the presence bacteraemia and presence of a permanent of prosthetic valves or CRM devices was pacemaker. Khatib and Sharma9 reviewed 100% (95% CI 99–100%). Our findings the results of 877 episodes of SAB at a support other recent studies challenging single centre in the USA and suggested that the convention that all patients with SAB echocardiography could be dispensable in must have echocardiography. Only 65% uncomplicated community-associated and of all episodes included for evaluation health care-associated SAB, and limited to were investigated with echocardiography; subsets with clinical findings of endocar- the rate of evaluation with TOE was 16%. ditis, persistent bacteraemia, intracardiac Strongly positive associations for IE were devices, secondary foci, and relapse. A found with CO-CA and CO-HA SAB, docu- retrospective study of SAB in a tertiary mented prolonged bacteraemia, injecting referral centre in the UK10 found 3 cases drug use, and the presence of prosthetic of IE in 113 CVL-associated episodes of valves or CRM devices. SAB, all of which had the risk factor of a The recommendation for obligatory inves- prosthetic valve or CRM device. An evalu- tigation with TOE for all SAB is based on ation of 2 cohorts of HO-HA SAB in Europe 11 a high prevalence of IE in SAB even when and the US found that TOE might be classical risk factors or manifestations of dispensable in patients who did not fulfil complicated infection are absent, increased any of 6 high risk criteria. sensitivity of TOE over TTE,16 and a high Finally, a recently published Australian relapse rate if IE was undiagnosed in SAB study with a high rate of TOE found a NPV and undertreated with short course (< 14 for IE of 100% for patients with SAB not days) antibiotics.17 TOE was shown to be a meeting any of 3 criteria: community onset, reasonable strategy for determining which prolonged bacteraemia, and intra-cardiac patients with uncomplicated CVL-associated prosthetic devices.23 The strength of this SAB could be treated with short-course study’s findings may have been further antibiotics.18 Where data are available,8–11,17 reinforced by a more complete analysis of rates of IE in intravascular catheter-asso- clinical details, including CVL associated. Of ciated SAB are noted between 3% and 16%, note, in the 50 patients meeting the Heriot with higher rates in studies where CVL criteria in our cohort, 2 were diagnosed and peripheral venous line-associated SAB with IE. is not analysed separately and in earlier Several limitations must be considered in studies, perhaps reflecting improving evaluating the validity and generalisability infection control practices. Fowler et al of these findings. Analysis was retrospective identified a population with intravascular in a single tertiary referral centre with catheter-associated SAB, with neither no blinding. There are multiple potential haemodialysis nor a permanent foreign sources for bias: follow-up blood cultures body, that they considered at low risk for are more likely to be performed when IE is complicated SAB.19 In three recent studies suspected, and the risk factors of persistent advocating for echocardiography in all bacteraemia and prosthetic heart valves are cases of SAB,20–22 full information about also part of the modified Duke’s criteria for patients with CVL-associated SAB and IE is diagnosing IE. Although the rate of echocar- not stated, including the presence of risk diography performance at 65% was higher factors for IE such as the presence of pros- than in rates of 37–59% in recent series11,20 thetic valves. it was still significantly less than 100% and
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 33 www.nzma.org.nz/journal ARTICLE
differences exist between patients with SAB reflect a degree of sampling bias as 73 receiving and not receiving echocardiog- patients with SAB were excluded as a TOE raphy. The percentage of episodes of SAB was not performed. By way of contrast, a having TOE performed was low at 16%. recent pooled analysis of five prospective This compares of rates of 12%10 to 59% in observational studies of SAB found a prev- a prospective study attempting to achieve alence of IE of 5.7% (UK) to 11.6% (US).26 100% evaluation with TOE.20 The sensi- Furthermore, even with a higher rate of tivity of TTE for detection of IE was 65%, TOE (39%), Heriot et al demonstrated IE lower than the 89% reported by Casella et in 56 of 532 patients with SAB (10.5%).23 al when good image quality was obtained,24 In order to avoid misclassification in but consistent with the findings of a recent the present study, patient records after Australian series of 483 patients with discharge were reviewed and 100 day suspected endocarditis investigated with TTE follow-up data on mortality and devel- and TOE which found a sensitivity of TTE of opment of endocarditis was available 68% for detecting native valve endocarditis.13 for 98.6% of episodes. No patients with SAB who did not have echocardiography Although the optimal timing of echocar- performed were re-admitted with IE within diography to maximise the sensitivity of 100 days. It is possible that episodes of IE detection of IE has not been established, may have been missed if sudden death we did note that in our series TTE was occurred due to IE without a diagnosis conducted at a median of 3 days, earlier having been made, or if clinically incon- than that of other series, and the optimal spicuous IE was undiagnosed but cured by a 5–7 day period recommended in a recent short course of antibiotics. We regarded the 25 review by Holland et al. This may have former scenario as likely to be infrequent, impacted on the sensitivity of TTE. Of note and the latter as of uncertain relevance. however, the 28 patients who were inves- Confirmation of these findings is required tigated with a second TTE did not receive a with prospective multicentre evaluation. change in diagnosis. In consultation with the departments of An important finding was the prev- infectious disease and cardiology, these alence of definite IE in this cohort of results are being used to improve the quality patients with SAB was at the lower end of delivery of care to patients with SAB by the range reported globally. This is likely ensuring that follow-up blood cultures are to reflect in part differing inclusion and done on all patients with SAB, and recom- exclusion criteria. For example, the prev- mending that echocardiography be deferred alence of IE in Fowler et al’s 1997 study until review or discussion with the infectious was 25%;16 however, this prevalence may disease or clinical microbiology service.
Competing interests: Nil Author information: Nicholas Gow, Department of Microbiology, Auckland City Hospital; Boris S. Lowe, Green Lane Cardiovascular Service, Auckland City Hospital; Joshua Freeman, Department of Microbiology, Auckland City Hospital; Sally Roberts, Department of Microbiology, Auckland City Hospital. Corresponding author: Nicholas Gow, Department of Microbiology, Auckland City Hospital [email protected] URL: http://www.nzma.org.nz/journal/read-the-journal/all-issues/2010-2019/ 2015/vol-128-no-1416/6560
REFERENCES: 1. Williamson D, Zhang J, 2. Thwaites G, Edgeworth J, 3. Turnidge J, Kotsanas D, Ritchie S, et al. Staphylo- Gkrania-Klotsas E, et al. Munckhof W, et al. on coccus aureus infections in Clinical management of behalf of the Australia New New Zealand, 2000–2011. Staphylococcus aureus Zealand Cooperative on Emerg Infect Dis. 2014; bacteraemia. Lancet Infect Outcomes in Staphylococcal 20(7):1157-1162. Dis. 2011; 11: 208-22. Sepsis. Staphylococcus
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 34 www.nzma.org.nz/journal ARTICLE
aureus bacteraemia: a cus aureus bacteraemia. sociated Staphylococcus major cause of mortality CID 2011; 53:1-9. aureus Bacteremia. CID. in Australia and New 12. Matulevicius S, Rohatgi A, 2005; 40:695-703. Zealand. Med J Aust. Das S, et al. Appropriate 20. Holden E, Bashir A, Das 2009; 191(7):368-373. use and clinical impact of I, et al. Staphylococcus 4. Fowler V, Miro J, Hoen transthoracic echocardiog- aureus bacteraemia in a B, et al. Staphylococcus raphy. JAMA Intern Med. UK tertiary referral centre: aureus endocarditis: a 2013; 173(17):1600-1607. ‘transoesophageal echo- consequence of medical 13. Barton T, Mottram P, cardiogram for all’ policy. J Antimicrob Chemother progress. JAMA. 2005; Stuart R, et al. Transtho- 2014; 69: 1960-1965. 293(24):3012-3021. racic echocardiography 5. Walls G, McBride S, is still useful in the initial 21. Van Hal S, Mathur G, Kelly Raymond R, et al. Infec- evaluation of patients J, et al. The role of trans- tive endocarditis in New with suspected infective thoracic echocardiography Zealand: data from the endocarditis: Evaluation of in excluding left sided International Collabo- a large cohort at a tertiary infective endocarditis in ration on Endocarditis referral center. Mayo Clin Staphylococcus aureus Prospective Cohort Study. Proc. 2014; 89(6):799-805. bacteraemia. Journ Infect 2005. 51:218-221. NZMJ. 2009; 127:38-51. 14. Chu V, Sexton D, Cabell C, et 6. Nishimura R, Otto C, Bonow al. Repeat Infective Endo- 22. Rasmussen RV, Høst U, R, et al. 2014 AHA/ACC carditis: Differentiating Arpi M, et al. Prevalence Guideline for the Manage- Relapse from Reinfection. of infective endocarditis in ment of Patients with Clin Infect Dis. 2005; patients with Staphylococ- Valvular Heart Disease. 41(3):406-409. cus aureus endocarditis. EJE 2011; 12: 414-20. JACC. 2014; 63: e57-e185. 15. Li J, Sexton D, Mick N, 7. Mitchell D and Howden B. et al. Proposed modi- 23. Heriot G, Yeoh J, Street A, et Diagnosis and management fications to the Duke al. Echocardiography has and Staphylococcus aureus Criteria for the diagnosis minimal yield and may not be warranted in Staphylo- bacteraemia. Internal of infective endocarditis. coccus aureus bacteraemia Medicine Journal. CID. 2000; 30: 633-638. without clinical risk factors 2005; 35: S17-S24. 16. Fowler V, Li J, Corey GR, for endocarditis. Eur 8. Pigrau C, Rodríguez D, et al. Role of echocardi- J Clin Microbiol Infect Planes A, et al. Manage- ography in evaluation of Dis. 2015; Feb 26 (epub ment of catheter-related patients with Staphylococ- ahead of print) staphylococcus aureus cus aureus bacteraemia: 24. Casella F, Rana B, bacteraemia: when experience in 103 patients. Casazza G, et al. The may sonographic study JACC 1997; 30:1072-8. potential impact of be unnecessary? Eur 17. Jernigan J and Farr B. contemporary transthorac- J Clin Microbiol Infect Short-Course Therapy of ic echocardiography on the Dis. 2003; 22:713-719. Catheter-related Staphylo- management of patients 9. Khatib R and Sharma coccus aureus Bacteremia: with native valve endocar- M. Echocardiography is A Meta-analysis. Ann Intern ditis: a comparison with dispensable in uncom- Med. 1993. 119:304-311. transoesophageal echocardi- plicated Staphylococcus 18. Rosen A, Fowler V, Corey ography. Echocardiography. aureus bacteraemia. Medi- G, et al. Cost-Effectiveness 2009;26(8):900-906. cine. 2013; 92:182-188. of Transesophageal 25. Holland T, Arnold C, 10. Joseph J, Meddows Echocardiography To Fowler V. Clinical Manage- T, Webster T, et al. Determine the Duration ment of Staphylococcus Prioritizing echocardiog- of Therapy for Intravas- aureus Bacteremia: A raphy in Staphylococcus cular Catheter–Associated Review. JAMA. 2014; aureus bacteraemia. J Staphylococcus aureus 312(13):1330-1341. Antimicrob Chemother. Bacteremia. Ann Intern 26. Kaasch A, Barlow G, 2013; 68:444-449. Med. 1999. 130:810-820. Edgeworth J, et al. Staphylo- 11. Kaasch AJ, Fowler V, Reig 19. Fowler V, Justice A, coccus aureus bloodstream S, et al. Use of a simple Moore C, et al. Risk infection: a pooled criteria set for guiding Factors For Hematoge- analysis of five prospective, echocardiography in nous Complications of observational studies. J nosocomial Staphylococ- Intravascular Catheter-As- Infect. 2014; 68(3):242-51.
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 35 www.nzma.org.nz/journal ARTICLE
HIV-associated tuberculosis in Auckland Christopher Luey, David Milne, Simon Briggs, Mark Thomas, Rupert Handy, Mitzi Nisbet.
ABSTRACT AIM: New Zealand has low rates of disease caused by to Mycobacterium tuberculosis (TB) and Human Immunodeficiency Virus (HIV). This study is the first to describe a New Zealand cohort of patients with HIV-associated TB. METHOD: We retrospectively reviewed the clinical records, laboratory data and chest radiographs of all patients who were diagnosed with HIV-associated TB and who commenced treatment for TB disease at Auckland City Hospital between January 1997 and July 2009. RESULTS: During the 12-and-a-half year study period, 40 patients were diagnosed with HIV-associated TB. The median age was 37 years and the median CD4 count was 130 cells/mm3. Only 2 patients were New Zealand born. Twenty-four (60%) patients had known HIV infection prior to their diagnosis of TB disease. Two patients with known HIV infection and positive tuberculin skin tests had not received treatment for latent tuberculosis infection (LTBI). Twenty-three (58%) patients received antiretroviral treatment during their TB treatment. There were 21 episodes of treatment interruption or immune reconstitution inflamma- tory syndrome. Three (8%) patients died. CONCLUSIONS: New Zealand continues to have a low incidence of HIV-associated TB. Early HIV diagnosis with universal screening and the treatment of LTBI in persons living with HIV infection is key to minimising the disease burden.
Introduction Methods Tuberculosis (TB) is the most common Study Population opportunistic infection and cause of death The infectious diseases department at in persons living with Human Immunode- Auckland City Hospital (ACH) is the tertiary ficiency Virus infection (PLWHIV) in the referral centre for all PLWHIV in the world, with the greatest burden of disease Auckland and Northland regions, serving in those living in sub-Saharan Africa.1 an adult population of approximately 1.1 HIV infection is the strongest risk factor million people.5 All adults (≥15 years old), for progression from latent tuberculosis with diagnosed HIV infection, who were infection (LTBI) to TB disease.2 notified to the Auckland or Northland New Zealand has a low incidence of Regional Public Health Services as having TB (7.8/100,000) and a low prevalence TB disease between 1 January 1997 and 30 of HIV infection (<1/1,000).3, 4 This study June 2009 were included if treatment for describes the demographics, clinical TB disease was initiated at ACH. PLWHIV features and outcomes of patients diag- diagnosed with or treated for LTBI were nosed with HIV-associated TB in Auckland, not included. New Zealand, between 1997 and 2009 as Data collection to define the cohort and identify potential We reviewed the clinical records of all preventative strategies. patients with HIV-associated TB to obtain
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 36 www.nzma.org.nz/journal ARTICLE
Table 1: Demographics and co-morbidities of patients with HIV-associated tuberculosis Number Characteristic (%) Male gender 26 (65%) Ethnicity African 20 (50%) Asian 17 (42%) New Zealand European 2 (5%) South American 1 (3%) Viral hepatitis co-infection 10 (25%) Hepatitis B 4 (10%) Hepatitis C 5 (12%) Hepatitis B and Hepatitis C 1 (3%) Diabetes mellitus 2 (5%) Steroid treatment at diagnosis of tuberculosis 2 (5%) History of intravenous drug use 4 (10%) History of alcohol abuse 3 (8%) Psychiatric diagnosis 3 (8%) HIV related co-morbidity† 3 (8%)
†Pancytopaenia (n = 1), peripheral neuropathy (n = 1), cardiomyopathy and dementia (n = 1). information on patient demographics, 6 patients identified each year without presenting features, radiographic findings, a trend in incidence over time. These 40 treatment and clinical outcomes. Patients patients represent 1.6% of the 2,429 persons were classified as having definite TB disease notified with TB disease and 3.5% of the if M. tuberculosis was cultured from a 1,133 PLWHIV cared for at ACH during the clinical specimen. Patients were classified audit period. Thirty-seven (92%) patients as having probable TB disease if there was had definite TB disease and 3 (8%) patients no positive culture of M. tuberculosis but had probable TB disease. strong clinical suspicion of TB disease, with Patient demographics and co-morbidities or without visualisation of acid-fast bacilli at initial assessment are outlined in Table 1. on microscopy, leading to treatment for TB The median age was 37 (range 21–63) years. disease. TB resistance testing during the study period was phenotypic only and did All but 2 (5%) patients were migrants from not include identification of gene mutations Africa (n = 20), Asia (n = 17) or South America (eg, rpoB gene). (n = 1). The remaining 2 patients were New Zealand-born Europeans, both with a recent All available chest radiographs were history of incarceration in Thailand (for 5 and reviewed and reported by a single respi- 10 years). The self-reported ethnicity of the ratory radiologist. Previously defined 1,133 PLWHIV under the care of ACH during criteria were used to classify chest radio- the audit period was European (n = 595), graph findings into those most consistent African (n = 226), Asian (n = 116), Māori (n = with primary disease, post-primary disease 85), Pacific Person (n = 46), South American/ (reactivation), atypical patterns, miliary TB, Caribbean (n = 16), Middle Eastern (n = 4) and minimal change or normal.6 not reported (n = 45). The study received approval from the Northern X Regional Ethics Committee. Previous tuberculosis history Five (13%) patients had previously Statistical methods received treatment for TB disease. One Results are presented as medians (range) patient received 20 days of treatment in or frequencies (percentages). Africa 5 months before diagnosis of HIV-as- sociated TB in Auckland. Two patients had Results completed a 6 month course of treatment Initial assessment in South America and Africa; 1 year and We identified 40 patients who had HIV- 5 years respectively before diagnosis associated TB diagnosed between 1 January of HIV-associated TB. Two patients had 1997 and 30 June 2009. There were 1 to completed 14 month treatment courses
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 37 www.nzma.org.nz/journal ARTICLE
in New Zealand centres. Of the patients 2 (5%) patients. Lymphadenopathy (n = treated in New Zealand, there were 19) was the most common physical exam- concerns regarding compliance in one and ination finding and was present in 2 the other was treated with a rifampicin patients who had none of the symptoms sparing regimen following an episode of mentioned above. One patient presented rifampicin associated hepatitis. with a chest wall mass and 2 patients were Twenty-four (60%) patients were known asymptomatic. One asymptomatic patient to be living with HIV prior to their diag- had a history of prior partial treatment nosis of TB disease. The median duration of for TB disease in Africa and was found to known HIV infection prior to the diagnosis have an abnormal chest radiograph on of TB disease was 50 (range 4–144) months immigration screening. The other asymp- in the 22 patients for whom this information tomatic patient, a contact of a patient with was available. Of the 24 patients known to pulmonary TB disease, was diagnosed have HIV infection, 7 (29%) had documented on the basis of a TST conversion and a results of either a tuberculin skin test (TST) chest radiograph showing mediastinal or a QuantiFERON-TB Gold assay prior to lymphadenopathy. their presentation with TB disease. Two had Distribution of tuberculosis positive TST results but had not received Thirty-two (80%) patients presented with LTBI treatment. The remaining 5 patients pulmonary TB disease (PTB) including had negative LTBI screening test results. 20 who also had extra-pulmonary Two of these 5 patients had a positive TST or involvement. Extra-pulmonary TB disease QuantiFERON-TB Gold during investigations (EPTB) with no identifiable PTB was seen in for TB disease. Previous treatment for LTBI 8 (20%) patients (Table 2). Overall, 28 (70%) was not documented to have been given to patients had extra-pulmonary involvement. any patient. Lymph nodes were the most common site of Clinical findings EPTB (19/28). The most common symptoms at presen- Radiography tation were fever (n = 29, 73%), cough Chest radiographs were available for (n = 26, 65%), weight loss (n = 22, 55%) review in 24 of 32 (75%) patients with and sweats (n = 19, 48%). One or more of PTB (Table 3). Five (21%) were normal or these symptoms were present in 35 (88%) had minimal change and one (4%) had patients. Haemoptysis was reported in an atypical pattern. Of the 5 patients with
Table 2: Distribution of tuberculosis disease Distribution of disease Number (%) Pulmonary 32 (80%) Table 3: Chest x-ray appearances in cases with pulmonary only 12 pulmonary tuberculosis (n = 24) pulmonary & extra-pulmonary 20 Chest radiograph pattern Number (%) Lymph node 16 Normal 4 (17%) Blood culture 5 Minimal change 1 (4%) Urine 5 Atypical 1 (4%) Pleural effusion 2 Miliary 2 (8%) Central nervous system 2 Primary tuberculosis pattern 7 (29%) Extra-pulmonary only 8 (20%) Consolidation 4 Lymph node 3 Enlarged thoracic lymph nodes 4 Gastrointestinal tract 2 Pleural effusion 2 Urine 1 Post-primary tuberculosis pattern 9 (38%) Prostate 1 Consolidation 7 Chest wall 1 Cavitation 3 Pleural effusion 1 Tuberculoma 2 Pericardial effusion 1 Endobronchial spread 3
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 38 www.nzma.org.nz/journal ARTICLE
normal or minimal changes on chest radio- hepatitis and multi-organ failure 12 days graphs, 3 were smear and culture positive after initiation of ART, reintroduction of ART from sputa and 2 were considered to have failed to achieve complete viral suppression probable TB disease. Cavitation was seen in and ART was subsequently deferred until 3 (12%) patients. after completion of treatment for TB disease. Microbiology Of the remaining 17 (53%) patients who did not start ART during TB treatment; 7 (22%) Thirty-seven (92%) patients had M. tuber- patients started ART after completion of culosis cultured from at least 1 clinical treatment for TB disease (median CD4 count specimen. One case of probable HIV asso- 240 (range 233–361) cells/mm3), 4 (12%) ciated TB was culture negative despite being patients either transferred to another centre sputum smear positive on 2 occasions. Of for ongoing care or were lost to further those with PTB, 28 of 32 (88%) had positive follow-up and 6 (19%) patients remained sputum cultures and 19 of 32 (59%) were ART naïve at the time this study. smear positive. Extrapulmonary specimens yielding >1 positive cultures for M. tubercu- Six of 23 (26%) patients receiving ART losis were lymph node biopsy (n = 9), urine during treatment for TB disease developed (n = 6), blood (n = 5), pleural aspirate (n = 4), immune reconstitution inflammatory gastrointestinal tract biopsy (n = 2), pleural syndrome (IRIS) that consisted of fever biopsy (n = 2), liver biopsy (n = 1), chest wall with new pulmonary infiltrates (n = 1) or biopsy (n = 1) and prostate biopsy (n = 1). increased pain/swelling at a site of known TB disease (n = 5). The median presenting Seven of 37 (19%) M. tuberculosis isolates CD4 count at the start of TB treatment were resistant to one or more first line for these patients was 54 (range 31–131) anti-tuberculous agents: rifampicin and cells/mm3. Two of these patients required isoniazid (n=1), rifampicin alone (n=2), treatment with prednisone including one in streptomycin alone (n=1), isoniazid and whom ART was also temporarily withheld. streptomycin (n=1), and rifampicin, etham- butol and streptomycin (n=1). Nineteen (48%) patients, all with CD4 counts < 200 cells/mm3, received trimetho- HIV status and management prim-sulphamethoxazole (TMP-SMX) Twenty-four (60%) patients were known prophylaxis during their treatment for to have HIV infection prior to their diag- TB disease, usually at a dose of 480mg nosis of TB disease. Of these, 8 were already daily. Five (26%) patients were intolerant established on antiretroviral treatment of TMP-SMX due to one or more of the (ART) with a median CD4 count of 315 (range following: rash (n = 4), pancytopaenia (n = 2) 64-466). The median CD4 count for the 16 or vomiting (n = 1). patients with known HIV not on ART was 142 (range 21-420). Tuberculosis treatment Treatment for TB disease was interrupted Sixteen (40%) patients were diagnosed in 8 (20%) patients due to one or more of the with HIV infection at the time of their following adverse events: hepatitis (n = 6), presentation with TB disease. The median severe rash (n = 2), optic neuropathy (n = 1), CD4 count was 54 (range 4–423). The median angioedema (n = 1) and severe abdominal duration of residence in New Zealand prior pain (n = 1). Three patients had more than to their diagnosis of TB disease was 14 one treatment related adverse event. Two (range 0–83) months. patients who developed treatment-related Thirty-two (80%) patients were ART naïve hepatitis had pre-existing viral hepatitis; 1 at the time they were diagnosed with TB was co-infected with hepatitis B and another disease. Fifteen (47%) started ART during with hepatitis B and hepatitis C. First line treatment for TB disease (median CD4 count treatment for TB disease was successfully 54 (range 2–418) cells/mm3), 10 in the first re-established in 3 patients (including 8 weeks of treatment and 5 after more than the patient with hepatitis B co-infection). 8 weeks of treatment. Two patients stopped Five (13%) patients with treatment related ART during treatment for TB disease. One adverse events received one or more second could not tolerate ART due to nausea and line anti-tuberculous agents; 3 because of vomiting and declined to start ART again. treatment related adverse events with first The other patient developed lactic acidosis, line anti-tuberculous agents and 2 because
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 39 www.nzma.org.nz/journal ARTICLE
of treatment related adverse events and In view of the likely predominance of susceptibility test results that demonstrated reactivation disease, the low rate (29%) of resistance to one or more first line anti-tu- retrospectively available LTBI screening berculous agents. results in known PLWHIV was concerning. Outcomes This may in part be explained by the lack of a centralised recording system for LTBI The median follow-up was 2.7 (range screening results for the majority of the 0–12.2) years. Twenty-eight (70%) patients study period, raising the possibility that were followed by the ACH infectious tests may have been performed but that diseases department until completion of the results of these tests were unable to be TB treatment and 1 patient (3%) continued obtained retrospectively. TST results for to receive treatment at the time of this example, the most commonly utilised LTBI study. Three (7%) patients died during screening test during the study period, TB treatment: 1 with fulminant PTB after were found only when documented in 7 days of TB treatment; 1 from B-cell handwritten clinic notes. All hospital and lymphoma after 7 months of TB treatment community TST and QuantiFERON-TB Gold and 1 after he declined all further test results are now recorded in a unified treatment. The remaining 8 (20%) patients electronic laboratory reporting system were transferred to another centre during which is easily interrogated. TB treatment. Also of concern was the finding that There were no cases of TB disease relapse. 2 PLWHIV with positive LTBI screens Four (10%) patients developed opportu- who did not receive LTBI treatment and nistic infections (OIs) after initiation of subsequently progressed to TB disease, TB treatment; Pneumocystis jiroveci pneu- represented missed opportunities to monia (n = 2), cerebral toxoplasmosis (n = prevent TB disease. It was not evident 1) and cryptococcal meningoencephalitis (n retrospectively whether LTBI treatment had = 1). Both P. jiroveci infections occurred in been offered. HIV infection is the greatest patients soon after initiating ART. Both of risk factor for reactivation of LTBI to TB these patients commenced ART 2 months disease.2 There is convincing evidence that after being diagnosed with TB. The other 2 LTBI treatment for PLWHIV significantly OIs occurred in patients who did not receive decreases their chances of developing TB ART, 15 months and 10 months after the disease. For PLWHIV, with a positive TST, start of TB treatment. treatment of LTBI was associated with a 0.38 relative risk of TB disease.7 Interna- Discussion tional and New Zealand guidelines now Our study demonstrates ongoing cases of advocate universal LTBI screening for HIV-associated TB over a 12-and-a-half-year all PLWHIV, and in the absence of strong period. New Zealand has a low incidence of contraindications or a suspicion of active TB disease (7.8/100,000) and a low prev- TB disease, all PLWHIV identified with alence of HIV infection (<1/1000).3, 4 All LTBI should be offered LTBI treatment.4,8,9 patients identified in our cohort either orig- In our cohort, 5 of 7 patients known to inated from countries with high rates of TB have HIV infection had negative Mantoux and HIV infection or had been incarcerated or QuantiFERON-TB Gold tests. Although in countries with high rates of TB and HIV this highlights the real possibility of false infection. The bulk of disease in our cohort negatives in PLWHIV, because the majority is more likely due to reactivation of LTBI of the benefit from preventive therapy rather than endemic cross-transmission. remains in Mantoux positive cases, we With the ongoing trend of globalisation would not advocate universal preventive and an annual intake of 750 United Nations therapy for PLWHIV from endemic coun- mandated refugees to New Zealand from tries with negative LTBI screening in the countries known to have high rates of absence of literature supporting benefit in 10 TB and HIV infection (including refugees this cohort. diagnosed with HIV infection), we expect to Early diagnosis of HIV infection continue to see patients with HIV-associated represents another important intervention TB disease in the future. for the prevention of TB disease. The risk of
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 40 www.nzma.org.nz/journal ARTICLE
LTBI reactivation increases with increasing mortality.13, 15, 16 In our cohort there were immune deficiency.11 The majority of 21 episodes of adverse treatment reactions patients in our cohort presented with (IRIS or an adverse reaction necessitating advanced HIV infection (median presenting interruption of TB treatment, ART or CD4 count 130 cells/mm3) including a TMP-SMX prophylaxis) and an 8% mortality significant proportion diagnosed with HIV rate. These rates are similar to the 54% infection only during their presentation adverse reaction rate and an 8.5% mortality with TB disease. For some, this presen- rate in a London cohort managed in the late tation was their first encounter with New 1990s in whom 54% received ART during Zealand health services, however this was their TB treatment course.16 Since the 1990s, not universally the case. Earlier detection of the treatment of HIV-associated TB has HIV infection could potentially have led to a undergone significant change. The evidence decrease in the risk of LTBI reactivation by for earlier introduction of ART is compelling virtue of earlier treatment of HIV infection, with large randomised control trials demon- thereby avoiding more advanced immu- strating clear mortality benefits when ART nosuppression, and also by detection and is introduced within the first 2 months of TB treatment of LTBI via the routine screening treatment.17, 18 This mortality benefit is most of all PLWHIV for LTBI. marked in the subgroup of patients with CD4 counts <50cells/mm3.18, 19 The benefits Diagnosis and management of HIV- of an earlier introduction of ART has come associated TB remains challenging. The with a cost; all three randomised control presentation of HIV-associated TB in our trials found high rates of adverse events cohort was similar to other previously with significant increases in the incidence reported cohorts with 35 of 40 presenting of IRIS with earlier ART.17-19 with WHO screening symptoms of fever, current cough, night sweats or weight loss.12-14 However, a significant proportion Conclusions of patients did not present with classical HIV-associated TB continues to be an symptoms. Although chest radiography is important problem in PLWHIV in New an important screening tool for PTB, it is Zealand. Reactivation disease was a signif- an insensitive one. Our study identified icant contributor to the burden of disease 5 cases of PTB where the presentation in our cohort. Earlier HIV diagnosis and chest radiograph was normal or showed treatment and the universal screening only minimal change. This is consistent and treatment of LTBI in PLWHIV remains with a previous study that found that 19% crucial to the control of both infections. of HIV-associated pulmonary TB cases Despite rapid advances in treatment had normal or minimal change chest guidelines for HIV-associated TB the radiographs.6 These findings emphasise morbidity and mortality associated with the importance of clinical vigilance and a HIV-associated TB remains high, further high index of suspicion for the detection of emphasising the importance of disease HIV-associated pulmonary TB. prevention. HIV-associated TB and its treatment is associated with a high rate of morbidity and
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 41 www.nzma.org.nz/journal ARTICLE
Competing interests: Nil Acknowledgements: The authors gratefully acknowledge the contribution of Helen Mills and Dr Cathy Pikholtz (Auckland Regional Public Health Service), Drs Jonathan Jarman and Loek Henneveld (Northland Regional Public Health Service) and Sue McAlister and Dr Nigel Dickson (AIDS Epidemiology Group). Author information: Christopher Luey, Infectious Diseases Physician,Infectious Diseases Department, Middle- more Hospital, Auckland; David Milne, Radiologist, Radiology Department, Auckland City Hospital, Auckland; Simon Briggs, Infectious Diseases Physician, Infectious Diseases De- partment, Auckland City Hospital, Auckland; Mark Thomas, Infectious Diseases Physician, Infectious Diseases Department, Auckland City Hospital, Auckland; Rupert Handy, Infectious Diseases Physician, Infectious Diseases Department, Auckland City Hospital, Auckland; Mitzi Nisbet, Infectious Diseases and Respiratory Physician, Infectious Diseases Department, Auck- land City Hospital, Auckland. Corresponding author: Christopher Edward Luey, General medical and Infectious Diseases Physician, Middlemore Hospital, 100 Hospital Road, Papatoetoe, Auckland, New Zealand 2025. [email protected] URL: http://www.nzma.org.nz/journal/read-the-journal/all-issues/2010-2019/ 2015/vol-128-no-1416/6561
REFERENCES: 1. WHO HIV/TB Facts 2014. of latent tuberculosis A. Clinical presentation Edition, cited. Available infection in HIV infected of tuberculosis and the from: http://www. persons. The Cochrane degree of immunodeficien- who.int/mediacentre/ database of systematic cy in patients with HIV factsheets/fs104/en/ reviews. 2010:CD000171. infection. Scandinavian 2. McShane H. Co-infection 8. Treatment of Tuberculosis journal of infectious with HIV and TB: double Guidelines. 4th Edition, diseases. 1999; 31:387-91. trouble. International cited.Available from: 13. Putong NM, Pitisuttithum journal of STD & AIDS. http://www.who.int/tb/ P, Supanaranond W, et al. 2005; 16:95-100; quiz 1. publications/tb_treatment- Mycobacterium tubercu- guidelines/en/index.html 3. UNAIDS Report on the losis infection among HIV/ Global AIDS Epidemic 9. Panel on Antiretroviral AIDS patients in Thailand: 2010. Edition, cited. Guidelines for Adults clinical manifestations and outcomes. The Southeast Available from: http://www. and Adolescents. Asian journal of tropical unaids.org/globalreport/ Guidelines for the use of antiretroviral agenst medicine and public Global_report.htm in HIV-1-infected adults health. 2002; 33:346-51. 4. Guidelines for tuber- and adolescents. 2011. 14. Raviglione MC, Narain JP, culosis control in New 10. Guidelines for intensified Kochi A. HIV-associated Zealand. 2010. tuberculosis case-finding tuberculosis in develop- 5. New Zealand Census. 2006. and isoniazid preventive ing countries: clinical 6. Greenberg SD, Frager D, therapy for people living features, diagnosis, and Suster B, Walker S, Stav- with HIV in resource-con- treatment. Bulletin of the ropoulos C, Rothpearl A. strained settings. 2011. World Health Organiza- Active pulmonary tubercu- 11. Williams BG, Dye C. tion. 1992; 70:515-26. losis in patients with AIDS: Antiretroviral drugs 15. Manosuthi W, Chottana- spectrum of radiographic for tuberculosis control pand S, Thongyen S, findings (including a in the era of HIV/AIDS. Chaovavanich A, Sung- normal appearance). Science. 2003; 301:1535-7. kanuparph S. Survival Radiology. 1994; 193:115-9. 12. Lado Lado FL, Barrio rate and risk factors of 7. Akolo C, Adetifa I, Shepperd Gomez E, Carballo Arceo E, mortality among HIV/ S, Volmink J. Treatment Cabarcos Ortiz de Barron tuberculosis-coinfected
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 42 www.nzma.org.nz/journal ARTICLE
patients with and without 17. Abdool Karim SS, Naidoo K, with tuberculosis. The New antiretroviral therapy. Grobler A, et al. Timing of England journal of medi- J Acquir Immune Defic initiation of antiretroviral cine. 2011; 365:1471-81. drugs during tuberculosis Syndr. 2006; 43:42-6. 19. Havlir DV, Kendall MA, therapy. The New England 16. Dean GL, Edwards SG, journal of medicine. Ive P, et al. Timing of Ives NJ, et al. Treatment of 2010; 362:697-706. antiretroviral therapy tuberculosis in HIV-infected 18. Blanc FX, Sok T, Laureillard for HIV-1 infection and persons in the era of highly D, et al. Earlier versus later tuberculosis. The New active antiretroviral ther- start of antiretroviral ther- England journal of medi- apy. AIDS. 2002; 16:75-83. apy in HIV-infected adults cine. 2011; 365:1482-91.
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 43 www.nzma.org.nz/journal ARTICLE
Efficacy of intralesional triamcinolone injections for benign refractory oesophageal strictures at Counties Manukau Health, New Zealand Yeri Ahn, Christin Coomarasamy, Ravinder Ogra
Abstract BACKGROUND AND AIMS: Refractory benign strictures of the oesophagus can present a significant clinical challenge and may require repeated attempts at dilatation. Intralesional triamcinolone injections have been used in a limited number of studies to aid in the dilatation of benign, refractory oesophageal stric- tures. It is hypothesised that intralesional steroid injection inhibits the scar formation, thereby reducing the need for dilatations. The purpose of this study was to assess if steroid injection reduces the need for serial endoscopic dilatations and aids in maintaining oesophageal patency. METHOD: We report a case series of 25 patients with refractory oesophageal strictures. The mean age was 75 years and range between 41–95 years. The etiology of strictures consisted of peptic (68%), anastomotic (4%), radiation induced (16%) and eosinophilic oesophagitis (12%). Majority (75%) were distal in location. Triamcinolone (40-80 mg) was injected via 25-gauge sclerotherapy catheter in a four quadrantic manner in aliquots of 0.5 ml each into the proximal end of the stricture and also into the stricture itself after dilata- tion. Dilatation was carried out with over the guidewire Savary-Gilliard or through the scope (CRE) Balloon dilators depending upon the preference of the endoscopist. Dilatations were continued every 4–6 weeks until asymptomatic and/or endoscopic resolution of stricture. Periodic Dilatation Index (PDI) was calcu- lated by the number of dilatations required over the duration of time in months. RESULTS: The number of dilatations reduced from mean of 3.12 to 1.41 in the peptic strictures but there was no decrease in the other groups. However, the triamcinolone injection resulted in reducing the periodic dilatation index in all groups except the eosinophilic oesophagitis. The rate ratio of PDI before and after intralesional triamcinolone injection use being 0.45 with 95% confidence interval [0.30 -0.68] (p=0.0005). Furthermore triamcinolone injections showed a trend to increase the maximal achieved diameter of the strictures. CONCLUSION: This study demonstrates the efficacy of triamcinolone intralesional steroids in reducing the requirement for repeated dilatations in refractory peptic strictures of oesophagus. Strictures related to eosinophilic oesophagitis failed to demonstrate similar efficacy.
In the last decade, intralesional steroid Introduction injections have become first line treatment Since Ashcraft and Holder first demon- in the management of refractory strictures, strated the effectiveness of intralesional especially in cases of oesophageal peptic corticosteroid injections in benign esoph- strictures.2–7 Studies in the management of ageal strictures using animal models in complex strictures, such as anastomotic8–10 1969,1 steroid injections have been utilised and radiation-induced3, 11 strictures, have as adjunctive aids to endoscopic dilatation also shown that steroid injections in combi- in gastrointestinal stricture management. nation with endoscopic dilatations have
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 44 www.nzma.org.nz/journal ARTICLE
largely reduced the need for repeated dila- alent of omeprazole and advised lifestyle tations and increased the intervention free anti-reflux measures. interval between dilatations. Fewer studies All patients underwent oesophageal have shown similar positive outcomes in dilatations at Counties Manukau Health managing strictures in pyloric stenosis and Gastroenterology service to relieve 12–15 Crohn’s. dysphagic symptoms. Before dilatations, Triamcinolone and other steroids have patients were sedated by intravenous historically been used in the treatment of administration of midazolam and fentanyl dermatological scars such as keloids and after topical lignocaine spray. Standard burns.16,17 On a microscopic level, similarly gastroscopes were used, however the tech- to scars, stricture formation is thought to nique of dilatation was variable between be inhibited by steroids, as it interferes endoscopists. with collagen synthesis and reduces the Savary-Gilliard and CRE (through the transcriptions of matrix protein genes such scope) balloon were the most commonly as fibronectin and procollagen, as well as used dilators. When intralesional steroids reducing synthesis of alpha2-macroglobulin, were used, dilatations were combined with an inhibitor of collagenase. triamcinolone acetonide (40mg/mL, diluted Hence steroid injections are thought to 1:1 with saline solution) injections in four further reduce the formation of strictures quadrants using a 23 gauge, 5mm long as it impairs fibrotic healing. sclerotherapy needle in aliquots of 0.5mL Currently there is no published data at the proximal margin of the stricture and in New Zealand to support the efficacy into the strictured segment. Most subse- of intralesional triamcinolone use in quent dilatation procedures were scheduled refractory oesophageal strictures. in 4–6 weeks until endoscopic resolution of strictures and improvement in patient’s The aim of this investigation was to symptoms, or earlier if patients returned objectively examine the use of intralesional with further dysphagia. triamcinolone injections in the endoscopic management of refractory oesophageal The primary investigator manually went strictures at Middlemore Hospital, Counties through databases (Endoscribe, Provation, Manukau Health, New Zealand, and to Concerto and hardcopy of clinical notes) to affirm its efficacy in improving clinical retrieve the site(s), number(s), nature and outcomes in patients with stricture disease. location of stricture(s) that were dilated, and recorded data for each patient. Each Patients and methods endoscopic procedure with and without triamcinolone injection for each patient The study was approved by the Northern for the relief of dysphagic symptoms was X Regional Ethics Commitee. The study examined and data extracted. The data patient population comprised of 25 regarding patient demographics and stricture patients (15 men, 10 women: mean age characteristics are available in Table 1. 75.08 [SD=14.78] years; range 41–95 years), with refractory oesophageal strictures of The total number of dilatations prior to differing etiologies treated between June receiving triamcinolone injection and the 2004 to January 2012. Seventeen patients total number of dilatations with triamcin- had peptic, 1 anastomotic, 4 radiation- olone injections until resolution of stricture induced and 3 eosinophilic oesophagitis or cessation of dilatations were recorded. related strictures. Refractory strictures Duration of endoscopic therapy was divided in this study were defined as anatomic into two: the total number of months from restriction due to fibrosis requiring more the first endoscopic dilatation for dysphagic than 3 sessions of dilatation to maintain symptoms until the first dilatation with lumen of at least 14 mm or inability to triamcinolone injection, and the total maintain a satisfactory luminal diameter number of months from the first dilatation for 4 weeks once the target diameter of 14 with triamcinolone injection until the endo- mm had been achieved.18 Prior to dilata- scopic dilatation therapy ceased. tions, all patients were receiving proton The maximal stricture diameter attained pump inhibitors averaging 40–80 mg equiv- with each endoscopic treatment with and
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 45 www.nzma.org.nz/journal ARTICLE
Table 1. Summary of statistics for patient demographic data and stricture characteristics. Anastomotic Radiation Other Total Characteristics Peptic n=17 n=1 n=4 (EO*) n=3 n=25 Age Mean [SD] 73.88 81.25 75.67 75.08 69[–] [16.01] [14.64] [11.93] [14.78]
Range 41–95 - 60–93 66–89 41–95
Gender (M:F) 14:3 0:1 1:3 0:3 15:10
Site of strictures Proximal - - 2 (7.1%) 3 (10.7%) 5 (17.9%)
Mid - - 1 (3.6%) 1 (3.6%) 2 (7.1%)
Distal 17 (60.7%) 1 (3.6%) 1 (3.6%) 2 (7.1%) 21 (75%)
Number of strictures One 17 (68.0%) 1 (4%) 4 (16.0%) 0 22 (88%)
Two - - - 3 (12%) 3 (12%)
Three - - - 0 0
Length of stricture
<3cm 13 1 2 2 18 >3cm 4 0 2 1 7 *EO: Eosinophilic Oesophagitis without triamcinolone injections was also Data regarding effects of intralesional recorded. Clinical records were searched triamcinolone injections can be seen in for any reported adverse events related to Tables 2 and 3. the endoscopic treatment Effectiveness of triamcinolone injections Results in improving clinical outcomes for patients Inferential analysis of data was carried with refractory oesophageal strictures out using SAS statistical software, version was thought to: i) reduce the number of 9.3. Poisson regression, with a logarithmic endoscopic dilatations required once triam- link function, was carried out on the cinolone injection therapy was commenced; primary outcome of PDI against time with ii) decrease the total duration of endoscopic patient ID as random effect to account therapy required; and iii) allow attainment for the over dispersion in the model. The of higher maximal stricture diameter once absolute rates with 95% confidence interval triamcinolone therapy was commenced. before and after triamcinolone injection use The periodic dilatation index (PDI) was used was 0.32 [0.22–0.47] and 0.15 [0.097–0.22] to objectively define efficacy of intralesional respectively. The raw data, as seen in Table steroid injections. Most patients received 2, also support this reduction in PDI after 80 mg Triamcinolone at each procedure. triamcinolone injections. Rate ratio of PDI PDI was defined by the number of dilata- before and after intralesional triamcinolone tions required over the duration of time injection use was 1.45 with 95% confi- in months.4 Not enough objective data dence interval [0.30 -0.68] and p-value of was available to elucidate symptom-free intervals between endoscopic dilatations. 0.0005. This shows a statistically significant The refractoriness of the strictures was reduction in PDI, and hence, the efficacy judged entirely from endoscopic appear- of triamcinolone injections in reducing the ances. Since the preliminary results of number of dilatations required over time in this series, we have started a prospective months. Subgroup analysis did not demon- database of all refractory strictures and strate a similar effect in strictures related to incorporated dysphagia scores to improve eosinophilic oesophagitis. objective assessment and also to reduce Furthermore, the raw data seen in Table 3 operator bias. show an increase in the maximum dilated
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 46 www.nzma.org.nz/journal ARTICLE
Table 2: Overall improvement in PDI following intra-lesional triamcinolone injections. Peptic Anastomotic Radiation Other (EO*) Total Variables n=17 n=1 n=4 n= 3 n=25
No of dilations Before After Before After Before After Before After Before After
Mean 3.12 1.41 1 3 3.25 5.25 2.67 3.0 3.0 2.28 ±[SD] [4.03] [1.23] [4.5] [7.23] [2.89] [3.46] [3.77] [3.26]
Periodic dilation Before After Before After Before After Before After Before After index(PDI) Mean 0.65 0.27 1 0.05 0.50 0.28 0.16 0.39 0.58 0.28
±[SD] [0.70] [0.31] [0.24] [0.08] [0.08] [0.53] [0.61] [0.30]
*EO: Eosinophilic Oesophagitis
Table 3: Overall improvement in maximum dilatation size (mm) achieved following intral-lesional triamcinolone injections Variables Peptic Anastomotic Radiation Other (EO*) Total n=17 n=1 induced n=4 n= 3 n=25
Maximum dilated Before After Before After Before After Before After Before After diameter (mm) Mean [SD] 16.12 19.29 17 17 15.75 18.5 16.0 17.67 16.08 18.88
[2.52] [1.31] [2.22] [1.73] [1.73] [0.58] [2.27] [1.42]
Range 10-20 17-20 - 13-18 17-20 15-18 17-18 10-20 17-20
*EO: Eosinophilic Oesophagitis
diameter of strictures after the introduction reducing the number of endoscopic dila- of triamcinolone injections. A linear tations required and decreased the total regression model was applied to investigate duration of endoscopic therapy in peptic, if improvement in maximal diameter of anastomotic and radiation induced stric- strictures impacted PDI for strictures. It was tures. Strictures related to Eosinophilic found not to be significant at the 5% level oesophagitis did not show any difference. even after adjusting for confounders such as Furthermore, triamcinolone injections age and location of stricture site (proximal, allowed higher maximal stricture diam- mid and distal) with a p-value of 0.17. This eters to be achieved once therapy was was thought to be primarily due to restricted commenced. This result was observed sample size. across strictures of different etiologies. All 25 patients included in this retrospective One limitation to our retrospective study study tolerated the procedures well, and no is potential operator bias affecting the data adverse events were documented at follow up collated. After successful interim results time. of treating refractory oesophageal stric- tures with triamcinolone injections, the Discussion refractory nature of some strictures were There are only a small number of studies anticipated by endoscopists depending in current literature that support the use of on the endoscopic appearance and the intralesional triamcinolone steroid injec- length of the stricture, and triamcinolone tions for the treatment of benign refractory injections were commenced earlier. This oesophageal strictures. Our study results, may have affected the dilatation intervals comprising 25 patients, have indicated that and some patients may have had antici- triamcinolone injections are efficacious in pated dilatations depending on endoscopic
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 47 www.nzma.org.nz/journal ARTICLE
appearances. Furthermore, as 3 different efficacy in relieving long-term dysphagic endoscopists carried out the dilatation symptoms (<20%).24, 27 We have used 3 types procedures, there may have been slight of stents for refractory strictures at our operator bias, although procedures were unit. Two patients received biodegradable largely standardised. stents, with successful resolution of Overall, this study adds to the growing stricture in one (50%). One Polyflex plastic evidence of the efficacy of triamcinolone stent (Migrated) and 2 self-expanding metal intralesional injections on benign refractory stents (Niti-S Taewoong Medical Korea), one oesophageal strictures with positive results migrating and the other successful. The use in favour of triamcinolone injection use of incisional therapies and argon plasma to improve patient outcomes in most coagulation is reported in small numbers etiologies, except eosinophilic oesopha- and has not gained widespread use. In gitis. The use in this condition needs to be view of the relative inefficacy and the high explored further. There were no adverse adverse events related to these newer events related to the use of Triamcinolone endoscopic therapies for these strictures, injection. This may encourage endoscopists the use of intralesional triamcinolone to use this therapy as adjunct to dilatation injections is a cheap and simple endoscopic at an earlier stage. method of managing refractory and In recent times alternative methods of difficult to treat strictures. The treatment treating refractory oesophageal strictures, did not show efficacy in the subgroup of such as removable stents, biodegradable eosinophilic oesophagitis and the paucity stents and experimental incisional ther- of numbers of anastomotic and radiation apies, have evolved. The use of removable induced strictures in our sample would or biodegradable stents is associated warrant further exploration of the use in with a high rate of adverse reactions and these situations. migration. Our study clearly indicates benefit for Non-randomised studies examining peptic strictures and should be adopted as the use of self-expanding plastic stents the first choice of treatment of such cases in the treatment of refractory benign and could be attempted in other etiologies. esophageal strictures have shown high Encouraged by these results, we enroll complication rates, including migration all patients with refractory oesophageal (22%–81%), chest pain (11%), bleeding strictures prospectively into a database (8%), and perforation (5.5%).25–27 The relief and also record the dysphagia scores and of dysphagic symptoms is sustained only have educated the endoscopists regarding in a minority of patients, and in a small the correct injection technique of triamcin- number of patients, tissue hyperplasia can olone injection and also have incorporated impact on the removability of stents. The the use of removable self-expanding metal use of self-expanding, fully-covered metal stents with antimigratory mechanism stents has been reported, however also [EndoMAXX Oesophageal stent (Merit yield disappointing results in regards to its Medical Endotek USA)].
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 48 www.nzma.org.nz/journal ARTICLE
Competing interests: Nil Ethics Approval: Ethics approved by Northern X Regional Ethics Committee. Approval code: NTX/08/68/EXP Acknowledgements: Authors of this paper would like to also acknowledge Professor Alain Vandal from the Counties Health Research office in his supervision with the statistical analysis of data. Author information: Yeri Ahn , General Medicine, Middlemore Hospital Counties Health Auckland; Christin Coomarasamy, Biostatistics Research Office, Counties Health Auckland; Ravinder Ogra, Gastroenterology, Middlemore Hospital Counties Health. Corresponding author: Ravinder Ogra, Gastroenterology, Middlemore Hospital Counties Health [email protected] URL: http://www.nzma.org.nz/journal/read-the-journal/all-issues/2010-2019/ 2015/vol-128-no-1416/6562
REFERENCES: 1. Sabanathan S, Salama Sahin B. Intralesional esophageal carcinoma: FD, Morgan WE. steroid injection in benign Increased success using Oesophageal intramural esophageal strictures submucosal corticoste- pseudodiverticulosis. resistant to bougie dilation. roid injection (Abstract Thorax 1985; 40: 849-857 J Gastroenterol Hepatol W1501). Gastrointest 2. Kirsch M, Blue M, Desai 2004; 19: 1388-1391 Endosc 2004; 59: P241 RK, Sivak MV Jr. Intrale- 6. Ramage JI Jr, Rumalla A, 9. Catalano MF, Kaul V, sional steroid injections Baron TH, Pochron NL, Pezanoski J, Guda N, for peptic esophageal Zinsmeister AR, Murray Geenen JE. Endoscopic strictures. Gastrointest JA, Norton ID, Diehl N, treatment of anastomotic Endosc 1991; 37: 180-182 Romero Y. A prospective, strictures following 3. Lee M, Kubik CM, Polha- randomized, double-blind, resection of esophageal mus CD, Brady CE 3rd, placebocontrolled trial of carcinoma (Abstract S1305). Kadakia SC. Preliminary endoscopic steroid injection Gastrointest Endosc experience with endoscopic therapy for recalcitrant 2007; 65: AB139 intralesional steroid esophageal peptic stric- 10. Miyashita M, Onda M, injection therapy for refrac- tures. Am J Gastroenterol Okawa K, Matsutani T, tory upper gastrointestinal 2005; 100: 2419-2425 Yoshiyuki T, Sasajima K, strictures. Gastrointest 7. Rupp T, Earle D, Ikenberry Kyono S, Yamashita K. Endosc 1995; 41: 598-601 S, Lumeng L, Lehman G. Endoscopic dexamethasone 4. Kochhar R, Ray JD, Sriram Randomized trial of Savary injection following balloon PV, Kumar S, Singh K. dilation with/without intral- dilatation of anastomotic Intralesional steroids esional steroids for benign stricture after esophago- augment the effects of gastroesophageal reflux gastrostomy. Am J Surg endoscopic dilation in strictures [abstract]. Gastro- 1997; 174: 442-444 corrosive esophageal intest Endosc 1995; 41: 357 11. Kochhar R, Makharia GK. strictures. Gastrointest 8. Catalano MF, George S, Usefulness of intralesional Endosc 1999; 49: 509-513 Thomas M, Geenen JE. triamcinolone in treatment 5. Altintas E, Kacar S, Tunc Endoscopic treatment of of benign esophageal B, Sezgin O, Parlak E, anastomotic strictures strictures. Gastrointest Altiparmak E, Saritas U, following resection of Endosc 2002; 56: 829-834
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 49 www.nzma.org.nz/journal ARTICLE
12. East JE, Brooker JC, Rutter triamcinolone acetonide of benign esophageal MD, Saunders BP. A pilot (kenalog). Immediate conditions. Gastrointest study of intrastricture and long-term results. Endosc. 2008;67:20–25. steroid versus placebo Scand J Plast Reconstr 24. Eloubeidi MA, Talreja JP, injection after balloon Surg 1977; 11: 169-172 Lopes TL, et al. Success dilatation of Crohn’s stric- 18. Kochman ML, McClave SA, and complications asso- tures. Clin Gastroenterol Boyce HW. The refrac- ciated with placement of Hepatol 2007; 5: 1065-1069 tory and the recurrent fully covered removable 13. Ramboer C, Verhamme esophageal stricture: a self-expandable metal M, Dhondt E, Huys S, definition. Gastrointest stents for benign esopha- Van Eygen K, Vermeire Endosc 2005; 62: 474-477. geal diseases. Gastrointest L. Endoscopic treatment 19. Zein NN, Greseth JM, Endosc. 2001;73:673–681. of stenosis in recurrent Perrault J. Endoscopic 25. Dua KS, Vleggaar FP, Crohn’s disease with intrale- sional steroid injec- Santharam R, et al. balloon dilation combined tions in the management Removable self-expand- with local corticosteroid of refractory esophageal ing plastic esophageal injection. Gastrointest strictures. Gastrointest stent as a continuous, Endosc 1995; 42: 252-255 Endosc 1995; 41:596–8. non-permanent dilator in 14. Singh VV, Draganov P, 20. Altintas E, Kacar S, Tunc treating refractory benign Valentine J. Efficacy and B, et al. Intralesional esophageal strictures: a safety of endoscopic steroid injection in benign prospective two- center balloon dilation of esophageal strictures study. Am J Gastroenterol. symptomatic upper and resistant to bougie dilation. 2008;103:2988–2994. lower gastrointestinal J Gastroenterol Hepatol. 26. Oh YS, Kochman ML, Crohn’s disease strictures. 2004;19:1388–1391 Ahmad NA, et al. Clin- J Clin Gastroenterol 21. Hordijk ML, Siersema ical outcomes after 2005; 39: 284-290 PD, Tilanus HW, et al. self- expanding plastic 15. Lavy A. Triamcinolone Electrocautery therapy for stent placement for improves outcome in refractory anastomotic refractory benign esoph- Crohn’s disease stric- strictures of the esopha- ageal strictures. Dig Dis tures. Dis Colon Rectum gus. Gastrointest Endosc. Sci. 2010;55:1344–1348. 1997; 40: 184-186 2006;63:157–163. 27. Canena JM, Liberato 16. Ketchum LD, Smith J, 22. Canhoto M, Arroja B, MJ, Rio-Tinto RA, et al. Robinson DW, Masters Silva F, et al. Needle-knife A comparison of the FW. The treatment of incisional treatment of temporary placement hypertrophic scar, keloid refractory esophagic of 3 different self-ex- and scar contracture by caustic stenosis. Endosc. panding stents for the triamcinolone acetonide. 2011;43(Suppl 2):E386. treatment of refractory Plast Reconstr Surg 23. Holm AN, de la Mora benign esophageal 1966; 38: 209-218 Levy JG, Gostout CJ,et al. strictures: a prospective 17. Kiil J. Keloids treated Self-expanding plastic multi-centre study. BMC with topical injections of stents in treatment Gastroenterol. 2012;12:70
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 50 www.nzma.org.nz/journal ARTICLE
Barriers to successful cessation among young late-onset smokers Hayley Guiney, Judy Li, Darren Walton
ABSTRACT AIMS: To understand the barriers to cessation among young late-onset smokers (young adults who started smoking daily after turning 18). Such information is crucial to the development of interventions aimed at reducing the high smoking prevalence among young adults. METHOD: The New Zealand Smoking Monitor is a fortnightly telephone survey of current smokers and recent quitters. This study focused on responses from a group of late-onset smokers aged 18 to 28 years (N = 111), who were temporarily (for 11 fortnights) added to the monitor. RESULTS: Most respondents had low nicotine dependence and were actively trying to quit (81% had made at least one attempt that lasted 24 hours or longer in the last year). One-half had high self-efficacy to quit and three-quarters did not intend to use cessation aids. Smoking was tightly linked to drinking alcohol and conferred social benefits (eg, 51% agreed “smoking helped me to socialise”). CONCLUSION: The tendency not to use cessation aids, strong links between smoking and drinking, and the social benefits of smoking may act as barriers to successful cessation among young late-onset smokers. Policies and interventions aimed at breaking associations between smoking, drinking and socialising (eg, smokefree bars) could be effective for this group.
obacco control research conceptualis- establishment of smoking behaviours.2,4–8 es smoking initiation not as a single For example, a recent US study found that event, but as a process that occurs 25% of those who were never smokers T 1,2 through a series of stages. While there before age 18 smoked their first whole may be variations in the stages individuals cigarette between the ages of 18 and 21 pass through, the ages at which they enter (representing the ‘trying smoking’ stage a particular stage, and the time spent in of initiation).8 In New Zealand, the consis- each stage, it is generally accepted that one tently high smoking prevalence among moves from never smoking (pre-contempla- New Zealand young adults (eg, 25.4% for 20 tion) through to contemplation, trying their to 24-year-olds in 2012/13 compared with first few cigarettes, experimenting with 17.6% in the general population)9 coupled smoking, smoking regularly but infrequent- with a low and declining smoking preva- ly, and then smoking at an established level lence among adolescents9,10 indicates that (daily or almost every day).1 significant initiation may be occurring after Traditional views of smoking initiation age 17. Recent longitudinal data support hold that the entire process from never to this contention, with 7% of never-smoking established smoking occurs almost exclu- 18 to 19-year-olds in New Zealand becoming sively in adolescence and consequently, regular smokers over the next four years preventive tobacco control initiatives (representing progression through all stages 7 have typically ignored young adults.3,4 of smoking initiation after turning 18). However, emerging evidence shows that Several possible reasons for young adult young adulthood (typically defined as 18 to smoking initiation (referred to in this 30-years-old) also represents a vulnerable article as ‘late-onset smoking’) have been time for the initiation, development, and put forward in the literature. First, there is
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 51 www.nzma.org.nz/journal ARTICLE
the increased risk associated with signif- While qualitative studies provide insight icant life changes, where transitions to new into the social contexts and functions of environments, activities, and peer groups smoking for young late-onset smokers in provide opportunities for the adoption of New Zealand, no studies have attempted new behaviours such as taking up smoking.2 to quantify the smoking- and cessation- The psychological and physiological effects related attitudes and behaviours of this of nicotine also provide a mechanism for group. To this end, the current research coping with the stress and anxiety that used data from the 2013 New Zealand might come with these transitions.2 Young Smoking Monitor (NZSM) to examine the adults are at risk of taking up smoking as attitudes and behaviours of young adults they face a number of these significant and who became established smokers only potentially stressful life changes, including after turning 18-years-old. Specifically, this transitioning to life outside the parental study assessed current smoking behaviour, home, interacting with new peer groups, nicotine dependence, self-efficacy to quit, entering the workforce, establishing iden- intentions and attempts to quit, intended tities, and engaging with new activities. At use of cessation aids, and social experiences the same time, young adults are earning with smoking, with a view to understanding greater disposable incomes11 and gaining the barriers to successful cessation in young greater access to tobacco (through the late-onset smokers. Understanding these ability to legally purchase cigarettes at age barriers, which are largely modifiable, is 18) and alcohol (including legal purchase important as they directly inform the design and admission to pubs and clubs). Finally, and targeting of tobacco initiatives aimed alluring tobacco pack designs and branding at reducing the high smoking prevalence offer young adults a ‘label’ with which to among New Zealand young adults. express their personality and identity and may therefore play a role in supporting and Method 12 maintaining smoking in that age group. Survey design Many theoretical reasons have been put Data were collected as part of the New forward to explain late-onset smoking, but Zealand Smoking Monitor (NZSM), which few studies have actually examined the is an on-going fortnightly survey of current phenomenon. Most that have done so have smokers and recent quit attempters’ smok- used varying measures of initiation (eg, ing-related attitudes and behaviours. The some used ‘trying a first cigarette’,8 while NZSM typically contains three sample others used ‘established smoking’7), were groups (non quit attempters, recent quit based in North America, and were focused attempters, and serious quit attempters), on either quantifying late-onset smoking but for this study the authors temporarily or determining the predictors of smoking (from July to November 2013) added a initiation in young adulthood (which fourth sample group: young adult (18 to include the use of alcohol, boredom, stress, 28-year-olds) smokers who took up daily and exposure to smoking by friends and smoking (ie, had become established family).4 smokers) after age 18. There is, however, one recent qualitative The method for the NZSM has been study that provides some insight into the described elsewhere.14 In brief, the NZSM contexts of smoking initiation for young uses a panel method where respondents can adults in New Zealand.13 The authors of participate in up to six fortnightly inter- that study found that young adults started views. Those who drop out of the survey smoking somewhat incidentally as a way (due to completing six interviews, being out of participating in social events and inter- of contact, or refusing to participate) are acting with peers on drinking occasions; replaced by new respondents. A commercial they did not necessarily make a conscious, research agency conducts the fieldwork considered decision to begin smoking. In using computer-assisted telephone inter- this way, drinking alcohol and the asso- viewing (CATI). Although NZSM respondents ciated social context play an important can participate in up to six interviews, this role in the initiation and development of paper reports on data collected from late- smoking behaviour in young adults.13 onset respondents’ first interview only.
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 52 www.nzma.org.nz/journal ARTICLE
The New Zealand Health and Disability many, if any, quit attempts lasting 24 hours Ethics Committee has classified the NZSM or longer they had made in the last 12 as a low-risk study; it is therefore exempted months (transformed to yes/no responses from ethics review.15 for analysis); whether they agreed with Recruitment the statement “I intend to quit in the next 3 months” (5-point agreement scale); how The fieldwork agency typically recruits likely it is (out of 100) they would succeed respondents for the NZSM from a nation- if they tried to quit smoking (used as an ally-representative omnibus survey (non indicator of self-efficacy to quit); and their quit attempters and recent quit attempters) intended use of specific cessation aids (yes/ or the New Zealand Quitline (serious quit no responses to specific aids). To assess the attempters). However, the low prevalence social context of smoking, respondents were of young late-onset smokers in the New asked to indicate their agreement (5-point Zealand population necessitated a different agreement scale) with five statements about recruitment procedure for this sample. their smoking-related experiences in the Recruitment contractors obtained contact last 2 weeks. details for potential late-onset respondents using a street intercept method, where Analysis recruiters approached young people in Analyses were conducted using STATA/ public places in various suburbs of six IC 13.1. Proportions were first calcu- New Zealand cities (Auckland, Hamilton, lated to assess the frequency of particular Wellington, Christchurch, Palmerston behaviours in the sample. For questions North, and Dunedin). Recruiters asked if with responses on a 5-point agreement scale, the person would like to participate in the ‘strongly agree’ and ‘agree’ responses were survey before screening for eligibility. combined to indicate overall agreement with particular statements. Respondents who Eligible respondents were aged between refused to answer a particular question, or 18 and 28 years, had started smoking daily said ‘don’t know’, were excluded from that (ie, had become established smokers) after particular analysis. the age of 18 years, and had smoked most Logistic regression analyses were then days in the last 30 days. We included only used to examine differences in responses established smokers in this study to ensure by demographic and smoking-related that respondents had relatively homoge- variables. For the majority of the analyses, neous smoking behaviour. Those who met the independent variables were age and the criteria and agreed to participate were gender as previous research indicates later contacted by the fieldwork agency these factors are important correlates of and re-screened for consent and eligibility smoking behaviour.9 Although previous before participating in the survey interview. research has also identified ethnicity as an Questionnaire important correlate of smoking behaviour,9 This study focused on smoking-related ethnicity could not be examined due to low behaviours, cessation-related behaviours, numbers of Māori (n = 8), Pacific (n = 9), and and the social context of smoking. To assess Asian (n = 16) respondents in the sample. smoking-related behaviours, respondents The dependent variable was the specific were asked how old they were when they smoking-related behaviour, cessation-related first tried a cigarette (respondents all behaviour, or experience with smoking. For became established smokers after turning the analyses investigating intention to quit, 18 years old, but could have smoked their nicotine dependence, self-efficacy, and recent first cigarette at any age) and the two ques- quit attempts were included as additional tions in the Heavy Smoking Index (HSI):16 independent variables in order to better ‘number of cigarettes per day’ (CPD number understand the predictors of quit intention of cigarettes smoked per day and time among young late-onset smokers. to first cigarette after waking. To assess Statistically significant (p < .05) results are cessation-related behaviours, respondents highlighted in the result tables, but where were asked: whether they had made a quit the odds ratios indicate potentially important attempt that lasted 24 hours or longer in non-significant trends, these are discussed in the last 3 months (yes/no response); how the text.
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 53 www.nzma.org.nz/journal ARTICLE
Table 1: Smoking-related behaviours among young late-onset smokers Percent n Age at first cigarette 13 or under 6 6 14–15 17 19 16–17 23 25 18–19 39 42 20–21 9 10 22–23 4 4 24-25 2 2 26 or over 1 1 Number of cigarettes smoked per day 1–5 41 45 6–10 42 46 11–20 15 17 21–30 1 1 31+ 1 1 Time to first cigarette after waking Within 5 mins 9 10 6 to 30 mins 21 23 31 to 60 mins 27 30 After 60 mins 43 48 Nicotine dependence (HSI score) Low dependence (0–1) 66 73 Medium dependence (2–4) 33 36 High dependence (5–6) 1 1
HSI = Heavy Smoking Index May not add to 100% exactly due to rounding
self-reported smoking behaviours by age or Results gender (p > .1 in all cases). Smoking-related Respondents behaviours are summarised in Table 1. The final sample included 111 current smokers (68 males; 103 non–Māori). Mean Cessation-related behaviours age was 22.5 years (SD = 2.6). Recent quit attempts Smoking-related behaviours Recent quit attempts were common, with 81% of respondents attempting to quit in Respondents in the current study were the last 12 months and 50% attempting to restricted to those who had become estab- quit in the last three months. The likelihood lished smokers only after turning 18; of making a quit attempt recently did not however, they could have tried their first vary by gender or age (p > .3 in all cases). cigarette at any age. More than one-half of respondents (54%) had tried their Intention to quit first cigarette after turning 18, with 18 to Nearly one-half of respondents (47%) 19-years-old being the most common age intended to quit in the next 3 months. for first trying a cigarette. The majority One-half (50%) reported high self-efficacy to (66%) of respondents had low nicotine quit (76% or greater chance of succeeding), dependence due to their relatively light 46% reported medium self-efficacy (26–74% smoking levels (83% smoked fewer than ten chance), and 4% reported low self-efficacy cigarettes per day) and long delays to their (0–25% chance). The only predictor of first cigarette after waking (43% smoked intention to quit was having made a recent their first cigarette more than 60 minutes quit attempt (in the past three or 12 months; after waking). There were no differences in Table 2).
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 54 www.nzma.org.nz/journal ARTICLE
Table 2: Factors associated with intention to quit in the next 3 months Independent variable Variable categories Gender Female Male % 47 47 0.98 OR (95% CI) Ref (0.45, 2.12) Age 18–20 21–24 25–28 % 40 44 59 1.19 2.18 OR (95% CI) Ref (0.48, 2.96) (0.76, 6.30) Self-efficacy to quit Med High % 48 45 1.11 OR (95% CI) Ref (0.51, 2.39) Nicotine dependence Low Med % 51 38 1.66 OR (95% CI) Ref (0.72, 3.81) Quit attempt in last 3 months Yes No % 62 31 3.59* OR (95% CI) Ref (1.62, 7.98) Quit attempt in last 12 months Yes No % 53 20 4.50* OR (95% CI) Ref (1.39, 14.58)
Note: Significant effects are bolded and marked with an asterisk; low self-efficacy and high nicotine dependence were not assessed due to a low number of respondents in those categories
Intended use of cessation aids that “in the last two weeks, there has been To assess intended use of cessation aids, an occasion where I smoked because I respondents were read a list of possible was drinking”. Smoking also appeared aids and asked whether or not they would to facilitate social interactions for many use each one. After reading the list, respon- respondents, with 51% agreeing that “in dents were also asked if they would try to the last two weeks, smoking made it easier quit “without any support”. Intended use of for me to socialise” and 43% agreeing cessation aids was low (see Table 3), with that “in the last two weeks, smoking has 73% of respondents saying they intended helped me meet new people”. Respondents to quit “without any support”. Aside from also experienced some social discomfort intending to quit without any support, the around smoking, with 37% agreeing that most common intended cessation aid was “in the last two weeks, there has been an occasion where I felt embarrassed when electronic cigarettes (50%), followed by the smoking” and 34% that “in the last two internet (46%) and nicotine replacement weeks, I have felt uncomfortable around therapy (NRT; 45%). others because I smoke.” Intended use of specific cessation aids did Rates of agreement with statements not vary by gender, although there were about the social context of smoking did not some effects of age (see Table 3). Specifi- change significantly with age or gender cally, 21 to 24-year-olds were less likely than (Table 4), except that older respondents 18 to 20-year-olds to intend to quit without (21 to 24-year-olds and 25 to 28-year-olds) any support, and more likely to intend to go were more likely than 18 to 20-year-olds to their GP or nurse. Both groups of older to agree that there had been an occasion respondents were less likely than 18 to in the last two weeks where they felt 20-year-olds to intend to use Quitline. embarrassed when smoking. There were, Social context of smoking however, some non-significant trends in the Smoking was closely linked to drinking raw percentages, such that older respon- alcohol, with 85% of respondents agreeing dents appeared to be less likely than 18
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 55 www.nzma.org.nz/journal ARTICLE
Table 3: Intended use of cessation products or services by gender and age Gender Age
Overall Female Male 18–20 21–24 25–28
% 73 74 72 86 63 76 Without any 1.12 0.27* 0.51 support OR (95%CI) Ref Ref (0.45,2.78) (0.08,0.92) (0.13,2.05) % 50 61 42 48 52 46 Electronic OR 2.16 1.16 0.92 cigarettes Ref Ref (95%CI) (0.97,4.84) (0.46,2.93) (0.32,2.65) % 46 48 44 45 43 52 Internet 1.14 0.92 1.33 OR (95%CI) Ref Ref (0.52,2.52) (0.37,2.33) (0.46,3.90) % 45 48 43 48 47 38 NRT 1.19 0.95 0.67 OR (95%CI) Ref Ref (0.54,2.62) (0.37,2.44) (0.23,2.01) % 38 33 40 59 35 19 Quitline 0.74 0.38* 0.17* OR (95%CI) Ref Ref (0.33,1.68) (0.15,0.96) (0.05,0.57) % 31 29 32 17 44 23 GP or nurse 0.86 3.73* 1.44 OR (95%CI) Ref Ref (0.37,2.05) (1.22,11.44) (0.38,5.43) % 31 33 30 38 29 28 Mobile app 1.19 0.65 0.64 OR (95%CI) Ref Ref (0.51,2.78) (0.25,1.73) (0.25,1.73) % 28 33 24 21 35 21 Cessation OR 1.51 1.96 0.96 medication Ref Ref (95%CI) (0.62,3.70) (0.66,5.80) (0.25,3.67)
Note: Significant effects are bolded and marked with an asterisk; multiple responses were allowed and therefore results do not add to 100%; NRT = nicotine replacement therapy
Table 4: Social experiences with smoking in the last two weeks by gender and age Gender Age
Female Male 18–20 21–24 25–28
Smoked because I % 81 87 87 87 78 was drinking 0.67 1.03 0.54 OR (95%CI) Ref Ref (0.24, 1.89) (0.28, 3.86) (0.13, 2.16) Smoking made it % 51 51 60 54 37 easier for me to 0.99 0.77 0.39 OR (95%CI) Ref Ref socialise (0.46, 2.12) (0.31, 1.91) (0.13, 1.14) Smoking helped me % 37 47 53 44 30 meet new people 0.67 0.70 0.37 OR (95%CI) Ref Ref (0.31, 1.45) (0.29, 1.71) (0.12, 1.10) Felt embarrassed % 40 35 17 41 52 when smoking 1.19 3.44* 5.38* OR (95%CI) Ref Ref (0.55, 2.63) (1.14, 10.36) (1.59, 18.26) Felt uncomfortable % 46 28 13 37 42 around others 2.25 3.85 4.64 OR (95%CI) Ref Ref because I smoke (0.87, 5.84) (0.77, 19.31) (0.85, 25.30)
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 56 www.nzma.org.nz/journal ARTICLE
to 20-year-olds to report experiencing the Among respondents in this study, 18 to 19 social benefits of smoking (eg, smoking years was the most common age for trying making it easier to socialise) and more a first cigarette (39% of respondents). In likely to report experiencing the social costs contrast, only 16% of respondents first tried (eg, feeling uncomfortable around others a cigarette after age 20. This indicates that, because you smoke). at least for those who became established smokers only after turning 18, the early Discussion stages of young adulthood (as opposed to the later stages) could be a key risk period The current study confirms that smoking for first trying cigarettes. The greater like- initiation, including trying one’s first few lihood that respondents in this study tried cigarettes, is not exclusive to adolescence. cigarettes in early, compared to later, young It also highlights important features adulthood is consistent with previous New of an under-studied group. The results Zealand research using a nationally-rep- characterise young late-onset smokers as resentative sample of young adults, which relatively light (albeit daily or almost daily) showed that smoking initiation (defined in smokers with low nicotine dependence, that study as going from never smoking to active cessation behaviour, high self- regular smoking) is not uncommon for 18 and efficacy to quit, and low intention to use 19-year-olds, but is unlikely after age 25.7 effective cessation aids. They also reveal aspects of the social context of smoking that Cessation behaviour may serve to reinforce smoking behaviours The young late-onset smokers in this and act against successful cessation study were active quit attempters (eg, 81% attempts in this age group. had tried to quit in the last 12 months), Evidence for late-onset smoking many (47%) of whom intended to quit in the near future. Most respondents (73%) Findings from the current study support said they would not use any cessation previous international3–5,8 and national support for future quit attempts, but inter- research7,13 showing that young adulthood estingly, many also said they would use is an important period in the initiation and electronic cigarettes (50%) or the internet development of smoking. Over one-half of (46%) to help them quit. While there is no respondents in the current study tried their obvious explanation why some of those first cigarette (representing the ‘trying ciga- who said they would quit ‘without any rettes’ stage of smoking initiation) only after support’ also said they might use one of the turning 18. Given the dearth of information other cessation aids, it is clear that using on late-onset smoking in New Zealand, this no support is the least effective quitting finding warrants further investigation into method,20 and that there is no consistent the contexts in which young adults try their evidence showing that young late-onset first cigarette. Although previous studies smokers’ preferred forms of support (elec- have investigated the context of first trying tronic cigarettes21 and the internet22) are cigarettes and the sources of tobacco supply effective at improving quit success. Thus, for adolescents,17–19 few have considered there appears to be a need to increase the same in young adults. Future research young adults’ use of cessation aids and to should seek to understand: where young have targeted services that appeal to, and late-onset smokers tried their first cigarette; are effective for, this group. Given respon- what they were doing at the time; and who dents’ low nicotine dependence and the or where they got the cigarette from. Such relatively low efficacy of NRT for people information would enable development of with low dependence,23 such services might policy initiatives that target the environ- focus on the non-chemical (ie, social and ments and contexts in which young adults behavioural) reinforcers of smoking. For enter the early phases of smoking initiation. example, providing young adults with prac- For example, if young adults predominantly tical strategies to deal with social pressure try their first cigarette when out socialising to smoke in certain contexts. at licensed venues, completely smokefree bar policies could be effective in reducing The social context of smoking opportunities for young adults to try a Previous research shows that drinking cigarette. alcohol and the associated social setting
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 57 www.nzma.org.nz/journal ARTICLE
plays an important role in the development Over time, successive tobacco control and maintenance of smoking behaviour initiatives in New Zealand have helped to in young adults24,25 by facilitating social denormalise smoking, thereby reducing interactions on drinking occasions and its social benefits. However, the current lowering judgments of risk perceptions.13,26 finding that smoking still confers social The current study supports the idea that benefits for young late-onset smokers indi- smoking and drinking alcohol are tightly cates that more could be done to further linked, with 85% of respondents agreeing denormalise smoking in social contexts: there had been an occasion in the last 2 a concept that is supported by previous weeks where they smoked because they research with young adult smokers in were drinking. In line with previous New Zealand.13,26 One way to achieve this recommendations in this paper and by could be to introduce legislation aimed at other tobacco control researchers,26 these further limiting the public areas in which findings suggest that interventions and people can smoke. Another possibility is policies aimed at de-coupling drinking to implement young adult-targeted mass and smoking (eg, entirely smokefree bars, media campaigns aimed at denormalising restaurants, and pubs) could be effective smoking in social settings. For example, in reducing among young adults the initi- Canada’s “Quit the Denial” campaign, which ation and development of smoking, as sought to challenge the acceptability of well as promoting successful cessation by ‘social smoking’ in young adults, or New preventing cues to relapse. Support for Zealand’s “Stop before you Start” campaign, this idea comes from an evaluation of the which aims to show how young adults get impact of the 2003 Smokefree Environ- drawn into smoking somewhat incidentally, ments Amendment Act, which showed a and that tobacco is not a normal consumer reduction in the proportion of smokers product due to its addictive properties and reporting that they smoked ‘more than serious health consequences. normal’ when they were at cafes, bars, and Limitations 27 nightclubs. The current study represents an Many of the respondents in the current important first step in understanding young study reported both social benefits (eg, late-onset smokers in New Zealand, but it smoking making it easier to socialise) and has a number of limitations. First, its gener- social costs of smoking (eg, feeling embar- alisability to all young late-onset smokers rassed around others when smoking). These in New Zealand is limited by its relatively findings align well with previous quali- small sample and low numbers of Māori tative research conducted with young social and Pacific respondents. The reason for low smokers in New Zealand, which highlighted numbers of Māori and Pacific respondents the conflict young adults feel between the in the current study is not clear, although it immediate social benefits of smoking in could reflect a very low prevalence of late- certain contexts and wider societal disap- onset smoking in those groups (note that proval of smoking.26 Interestingly, we the prevalence of late-onset smoking among found some evidence that the likelihood of Māori and Pacific people is not currently experiencing the social costs of smoking known). It could also indicate that the increased with age, while the likelihood of recruitment procedure favoured non-Māori experiencing the social benefits decreased and non-Pacific people, despite recruitment (note however that although the odds ratios occurring in a number of different public were large, not all findings were statisti- places around New Zealand. Future cally significant). Further research with studies should include a larger and more a larger sample is needed to detect these representative sample that would enable effects and understand why they occur. If examination of late-onset smoking in Māori the social benefits of smoking do diminish and Pacific sub-groups. The small sample in throughout young adulthood, initiatives general may also account for the inability aimed at reducing the social reinforcement to detect differences between groups where of smoking should specifically target early the raw percentages suggested there may be young adulthood. a trend in the data.
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 58 www.nzma.org.nz/journal ARTICLE
A second limitation is that this study does occur in young adulthood. They also point not provide insight into the smoking- and to several factors that could contribute to cessation-related attitudes of those who had unsuccessful cessation attempts among started on the smoking initiation pathway, young late-onset smokers, despite their low but had not yet progressed to established nicotine dependence (a known predictor smoking (eg, still in the experimentation of quit success24) and active cessation phase), as the sample was restricted to behaviour. In particular, the tendency not established smokers. Third, respondents to use effective cessation aids, the strong were both young adults and late-onset association between smoking and drinking smokers, meaning that we could not alcohol, and the social benefits that smoking separate the features of late-onset smoking confers are all likely to reinforce smoking in from those of young adult smoking. Finally, this age group. the current study used a relatively simple In line with previous research, the measure of self-efficacy to quit (perceived current findings point to a number of initia- likelihood of success). More comprehensive tives that could contribute to a reduction in assessments of the construct may prove the high smoking prevalence among young more informative in this regard. adults in New Zealand. First, interventions Despite these limitations, the current that specifically target this crucial phase of findings provide information that can be development, during which life transitions used to develop effective interventions and changes in social networks increase the aimed at preventing the development risk for progressing along the smoking initi- 2 of a lifelong smoking addiction. Their ation continuum , are needed. For example, targeting young adults (as opposed to consistency with previous qualitative focusing on youth or adults generally) with research with young adult smokers in mass media campaigns aimed at denor- New Zealand13,26,28 also suggests that the malising smoking is likely to be important behaviours and experiences reported here for reducing initiation and encouraging reflect the early stages of a young person’s cessation in that age group.3,4 Cessation smoking career. services and aids that are attractive and effective for young adults are also needed. Conclusion Finally, policy changes aimed at breaking The current study provides useful insight the associations between smoking, into the ‘smoking lives’ of young late-onset drinking, and socialising (eg, smokefree smokers in New Zealand. The findings restaurant, bar, and street policies) should support the idea that smoking initiation can be considered.
Competing interests: Nil Acknowledgements: The authors thank Danny Tu for preparing the data for analysis and Karen McBride-Henry for her comments on the draft manuscript Author information: Hayley Guiney, Researcher, Research and Evaluation Unit, Health Promotion Agency, Wel- lington; Judy Li, Senior Researcher, Research and Evaluation Unit, Health Promotion Agency, Wellington; Darren Walton, Manager, Research and Evaluation Unit, Health Promotion Agency, Wellington. Corresponding author: Hayley Guiney, Research and Evaluation Unit, Health Promotion Agency, PO Box 2142 Wellington 6140. [email protected] URL: http://www.nzma.org.nz/journal/read-the-journal/all-issues/2010-2019/ 2015/vol-128-no-1416/6563
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 59 www.nzma.org.nz/journal ARTICLE
REFERENCES: 1. Mayhew KP, Flay BR, Cohort. J Adolesc Robinson J. Measuring Mott JA. Stages in the Health. 2012;51(5):497- the heaviness of smoking: development of adolescent 502. doi:10.1016/j. using self-reported time to smoking. Drug Alcohol jadohealth.2012.02.017. the first cigarette of the day Depend. 2000;59, Supple- 9. Ministry of Health. Tobacco and number of cigarettes ment 1:61-81. doi:10.1016/ Use 2012/13: New Zealand smoked per day. Br J S0376-8716(99)00165-9. Health Survey. Wellington: Addict. 1989;84(7):791-799. 2. Ling PM, Glantz SA. Why Ministry of Health; 2014. 17. Presti DE, Ary DV, Lichten- and how the tobacco 10. Action on Smoking stein E. The context of industry sells cigarettes smoking initiation and and Health. Factsheet: to young adults: evidence maintenance: Findings Youth Smoking in New from industry documents. J from interviews with Zealand. http://www. Inf. 2002;92(6). http://ajph. youths. J Subst Abuse. ash.org.nz/wp-content/ aphapublications.org/doi/ 1992;4(1):35-45.doi:10.1016/0899- uploads/2014/08/2013- pdf/10.2105/AJPH.92.6.908. 3289(92)90026-T. Youth-smoking-in-New- Accessed June 26, 2013. Zealand.pdf. Accessed 18. Eathorne A, White J. 3. Hammond D. Smoking October 1, 2014. Offer and Acceptance of behaviour among young Tobacco among Young 11. Muir K, Mullan K, Powell adults: beyond youth People. Wellington: A, Flaxman S, Thompson prevention. Tob Control. Health Promotion Agency D, Griffiths M. State of 2005;14(3):181-185. Research and Evaluation Australia’s Young People: doi:10.1136/tc.2004.009621. Unit; 2014. http://www.hpa. A Report on the Social, org.nz/sites/default/files/ 4. Freedman K, Nelson N, Economic, Health and Offer%20and%20Accep- Feldman L. Smoking Family Lives of Young tance%20of%20tobacco%20 Initiation Among Young People. Australian Govern- among%20young%20 Adults in the United States ment: Office for Youth and Canada, 1998-2010: peple%20%282%29.pdf. 12. Hoek J, Gendall P, Gifford A Systematic Review. 19. Marsh L, Dawson A, McGee H, et al. Tobacco branding, Prev Chronic Dis. 2011. R. “When you’re desperate plain packaging, pictorial doi:10.5888/pcd9.110037. you’ll ask anybody”: young warnings, and symbolic 5. Tjora T, Hetland J, Aarø people’s social sources of consumption. Qual Health LE, Wold B, Øverland S. tobacco. Aust N Z J Public Res. 2012;22(5):630-639. Late-onset smokers: how Health. 2013;37(2):155-161. many, and associations 13. Gray RJ, Hoek J, Edwards doi:10.1111/1753-6405.12033. with health behaviours R. A qualitative analysis 20. Lancaster T, Stead L, Silagy and socioeconomic status. of “informed choice” C, Sowden A. Regular Scand J Public Health. among young adult review: Effectiveness of 2012;40(6):537-543. doi: smokers. Tob Control. interventions to help people 10.1177/1403494812454233. 2014:tobaccocontrol - stop smoking: findings 2014-051793. doi:10.1136/ 6. O’Loughlin JL, Dugas EN, from the Cochrane Library. tobaccocontrol-2014-051793. O’Loughlin EK, Karp I, BMJ. 2000;321(7257):355. Sylvestre M-P. Incidence 14. Li J, Newcombe R, 21. World Health Organization. and Determinants of Ciga- Walton D. The use of Electronic Nicotine Delivery rette Smoking Initiation and attitudes towards Systems: Report by WHO.; in Young Adults. J Adolesc electronic cigarettes, and 2014. http://apps.who.int/ Health. 2014;54(1):26-32. self-reported exposure gb/fctc/PDF/cop6/FCTC_ to advertising and the 7. Edwards R, Carter K, Peace COP6_10-en.pdf?ua=1. product in general. Aust N J, Blakely T. An examination 22. Walters ST, Wright JA, Z J Public Health. in press. of smoking initiation rates Shegog R. A review of by age: results from a large 15. Health and Disability Ethics computer and Inter- longitudinal study in New Committees. Standard net-based interventions for Zealand. Aust N Z J Public operating procedures smoking behavior. Addict Health. 2013;37(6):516-519. for health and disability Behav. 2006;31(2):264- doi:10.1111/1753-6405.12105. ethics committies. 2012. 277. doi:10.1016/j. 8. Bernat DH, Klein EG, http://ethics.health.govt. addbeh.2005.05.002. Forster JL. Smoking nz/operating-procedures. 23. Tang JL, Law M, Wald Initiation During Young Accessed June 28, 2013. N. How effective is Adulthood: A Longitudinal 16. Heatherton TF, Kozlowski nicotine replacement Study of a Population-Based LT, Frecker RC, Rickert W, therapy in helping
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 60 www.nzma.org.nz/journal ARTICLE
people to stop smoking? role of lifestyle, attitudes/ evaluation of the impact of BMJ. 1994;308(6920):21- beliefs, demographics, a new national smokefree 26. doi:10.1136/ and exposure to anti-to- law in New Zealand. Tob bmj.308.6920.21. bacco media messaging. Control. 2008;17(1):e2-e2. 24. Gilpin EA, White VM, Drug Alcohol Depend. doi:10.1136/tc.2007.020347. Pierce JP. What fraction of 2013;130(1–3):115- 28. McCool J, Hoek J, Edwards young adults are at risk for 121. doi:10.1016/j. R, Thomson G, Gifford future smoking, and who drugalcdep.2012.10.019. H. Crossing the smoking are they? Nicotine Tob Res 26. Hoek J, Maubach N, divide for young adults: Off J Soc Res Nicotine Tob. Stevenson R, Gendall P, expressions of stigma 2005;7(5):747-759. Edwards R. Social smokers’ and identity among doi:10.1080/ management of conflicted smokers and nonsmok- 14622200500259796. identities. Tob Control. ers. Nicotine Tob Res 25. Dietz NA, Sly DF, Lee DJ, 2013;22(4):261-265. Off J Soc Res Nicotine Arheart KL, McClure LA. 27. Edwards R, Thomson Tob. 2013;15(2):552-556. Correlates of smoking G, Wilson N, et al. After doi:10.1093/ntr/nts136. among young adults: The the smoke has cleared:
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 61 www.nzma.org.nz/journal ARTICLE
Karyotypes, confined blood chimerism, and confusion: a case of genetic sex mislabelling and its potential consequences Aarthi Ravishankar, José G B Derraik, Sarah Mathai, Wayne S Cutfield, Paul L Hofman
ABSTRACT Disorders of sex development (DSD) encompass a range of conditions, and the management of infants with DSD can be extremely complex. However, the misdiagnosis of a normal infant as a case of DSD may lead to unfortunate long-term consequences for the individual and the family. We report a case of confined blood chimerism masking as 46 XY gonadal dysgenesis in a female from a twin pair with discordant genders, which led to incorrect sex determination at birth. The potentially serious consequences of a wrong DSD diagnosis are discussed, including the removal of normal ovaries. This case emphasises the importance of confirming a blood karyotype where there is discordance with the clinical phenotype and, where possible, identifying whether functional gonadal tissue is present.
at Starship Children’s Hospital (Auckland, Introduction New Zealand). The child was born at term Disorders of sex development (DSD) weighing 3070 g. The twin pregnancy encompass a range of conditions, but they involved complications, with an anatomy have been defined as congenital conditions scan at 18 weeks of gestation showing where the development of chromosomal, dysmorphic features in one twin, including gonadal or anatomical sex is atypical.2 The a nasal malformation and severe hyper- management of infants with DSD can be telorism. Amniocentesis was subsequently extremely complex, requiring the long-term performed, revealing a 46 XY karyotype for involvement of a multidisciplinary team both fetuses. No information was recorded working alongside the family.7 However, in as to whether this was a monochorionic or some cases a normal infant may be wrongly dichorionic twin pregnancy. labelled as having DSD, which may lead to At birth, Twin I presented as a phenotyp- unfortunate life-long consequences for the ically normal boy, while Twin II appeared individual and the family. Here, we present as a phenotypically normal female. a case of confined blood chimerism masking Examination of the female twin showed as 46 XY gonadal dysgenesis, which led to entirely normal female external geni- incorrect sex determination at birth. We talia. In particular, there was no evidence discuss the potentially serious ramifications of virilisation, with no labial fusion or of a wrong DSD diagnosis, including the cliteromegaly. No gonads were palpable. surgical removal of normal ovaries. An urgent blood lymphocyte karyotype confirmed a 46 XY karyotype. There was no Case Report evidence of mosaicism. A newborn twin (Twin II) was referred to Further studies were performed to the Department of Paediatric Endocrinology examine testicular and hypothalamic-
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 62 www.nzma.org.nz/journal ARTICLE
Table 1: Biochemical parameters in the female twin at age 2 months. Results Normal range Sodium (mmol/l) 136 133–146
Potassium (mmol/l) 4.8 3.7–5.2
Cholesterol (mmol/l) 3.7 <5.0
17-OHP (nmol/l) 2.0 <4.0
DHEAS (μ mol/l) <0.04 0.06–6.70
Renin (mU/l) 287 <780
ACTH (pmol/l) 11.9 2.0–11.0 17-OHP, 17-hydroxyprogesterone; ACTH,adrenocorticotropic hormone; DHEAS, dehydroepiandrosterone sulphate
Table 2: Results from stimulation tests in the female twin at ages 2 months and 9 years. The normal ranges are provided in square brackets. Age Stimulation Parameter Pre– Post– test stimulation stimulation 2 months hCG Testosterone (nmol/l) 0.4 [2-8] 0.3 [>8.0]
GnRH LH (IU/l) <0.8 1.8#
FSH (IU/l) 10.8 42.1#
9 years GnRH LH (IU/l) <0.8 [<2.0] 33.5*
FSH (IU/l) 1.9 [<2.0] 19.2* FSH, follicle stimulating hormone; GnRH, gonadotropin-releasing hormone; hCG, human chorionic gonadotropin; LH, luteinizing hormone. # From approximately 1 week to 6 months of age, GnRH stimulation tests havelead to pubertal responses (>5.0 IU/l).from approximately 1 week to 6 months of age. * Normal prepubertal baseline values usually are <5.0 IU/l, with post-stimulation LH levels ≥10.0 IU/l providing strong evidence of precocious puberty in females. pituitary function in Twin II. Assessment therefore be infertile), the child would of adrenal androgens (Table 1) excluded need pubertal induction and subsequently congenital adrenal hyperplasia. A human life-long sex steroid replacement. Early chorionic gonadotropin (hCG) stimulation childhood surgery to remove the presumed test confirmed the lack of functional testis, streak gonads was recommended, due with low basal testosterone concentrations to the risk of malignancy. However, the that did not rise with hCG stimulation (Table parents decided against gonadectomy and 2). A gonadotropin-releasing hormone asked for no further follow-up until it was (GnRH) stimulation test demonstrated a time for pubertal induction. normal follicle stimulating hormone (FSH) At the age of nine years, Twin II repre- rise but a blunted luteinising hormone sented with a six-month history of breast (LH) peak (Table 2). This was interpreted development. On examination substantial as likely to be normal, although a mild Tanner Stage 3 breast development and defect in hypothalamic-pituitary function Tanner Stage 2 pubic hair were noted. could not be excluded. Ultrasound scans The potential of an oestrogen-secreting demonstrated the presence of a uterus and gonadoblastoma was considered a possible fallopian tubes, but gonads were not visible. cause for this unexpected feminisation. The presence of a uterus in an individual Tumour markers were negative. Blood with a 46 XY karyotype is consistent with a samples were taken for a repeat chro- failure of testes development, and a diag- mosomal analysis, confirming a 46 XY nosis of XY gonadal dysgenesis was made. karyotype. Fluorescent in situ hybridi- The family were counselled that raising the sation (FISH) with the SRY probe also child as female was appropriate, and as the yielded positive results. A GnRH stimu- child had no functioning gonads (and would lation test showed findings consistent with
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 63 www.nzma.org.nz/journal ARTICLE
puberty (Table 2), and the bone age was of 8% amongst twin pairs and 21% in two years advanced. An MRI scan revealed triplets.9 However, very few cases have the presence of a normal vagina and a been reported in the literature, particularly normal-sized anteverted uterus with a thin regarding confined blood chimerism in endometrium. Unexpectedly, both ovaries dizygotic twins of opposite genders.6,8 were also identified and reported to have In the present case, the occurrence of normal morphology, with 2–3 simple cysts confined blood chimerism in Twin II led to noted bilaterally. the misdiagnosis of XY gonadal dysgenesis. Twin confined blood chimerism was Indeed, this misdiagnosis was felt to be so considered as a possible cause for these secure that gonadectomy in infancy was findings, and skin fibroblast analysis in recommended to the parents. Tragically, the Twin II was subsequently performed. This surgical removal of normal ovaries in such demonstrated a 46 XX karyotype confirming a situation has been reported previously, she was female. Lymphocyte DNA analysis with permanent infertility and life-long sex demonstrated her blood genotype was steroid requirements in adult life.8 Fortu- identical to her twin brother, indicating that nately, in this case the parents decided confined blood chimerism had occurred, against the recommended procedure. presumably due to a unilateral blood trans- A fundamental learning point from this fusion from her twin brother. case is the need to clearly define whether functioning gonadal tissue exists. When this Discussion child initially presented, an inhibin B assay was unavailable. Inhibin B is found at low This case highlights the importance but measurable levels in infant girls, and of fully investigating patients where the might have resulted in the accurate recog- phenotype is discordant with the karyotype. nition of functioning ovarian tissue, thus The assumption that blood lymphocytes avoiding the misdiagnosis of XY gonadal will provide a karyotype that reflects all dysgenesis.5 We would recommend that body tissues is not always correct. Mosaic in assessing gonadal function, stimulation and chimeric karyotypes are well docu- tests (such as using luteinising hormone or mented, are not always apparent on a blood hCG) as well assessment of inhibin B levels karyotype alone, and are not uncommon. should be used. Where possible, pelvic A chimera is an organism that contains MRI scans and/or laparoscopy should also 4,8 cells from more than one distinct zygote. be performed, as pelvic ultrasound scans This differs from mosaicism in which genet- are less sensitive, especially at identifying ically different cell lines originate from a gonadal tissue. single zygote lineage.3,6 In the majority of This case emphasises the importance cases, chimerism is seen after allogeneic of confirming a blood karyotype where bone marrow transplantation, in which there is discordance with the clinical bone marrow-derived cells belonging to the phenotype. In particular, where XY gonadal recipient are eradicated and substituted dysgenesis is being considered, assess- 1 with donor cells. ments of ovarian function are mandatory. However, there are cases reporting If there is suspicion of a chimera, fibro- confined blood chimerism occurring in blast chromosomal analysis in addition dizygotic twins. This is thought to occur via to a blood lymphocyte karyotype should an exchange of haematopoietic stem cells be performed. Notably, confined blood in utero, via inter-twin placental vascular chimerism in twins is not uncommon, espe- anastomoses.3,6 It has been suggested that cially when they are monochorionic, and confined blood chimerism in dizygotic this diagnosis should be considered in DSD twins has a higher incidence than previ- cases where there is a twin pregnancy and ously thought, with an occurrence rate discordant twin genders.
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 64 www.nzma.org.nz/journal ARTICLE
Competing interests: Nil Author information: Aarthi Ravishankar, Division of Imaging Sciences and Biomedical Engineering, King’s Col- lege London, London, England; José G B Derraik, Liggins Institute, University of Auckland, Auckland, New Zealand; Sarah Mathai, Liggins Institute, University of Auckland, Auckland, New Zealand; Wayne S Cutfield, Liggins Institute, University of Auckland, Auckland, New Zealand and Department of Paediatric Endocrinology, Starship Children’s Hospital, Auck- land, New Zealand; Paul L Hofman, Liggins Institute, University of Auckland, Auckland, New Zealand and Department of Paediatric Endocrinology, Starship Children’s Hospital, Auck- land, New Zealand. Corresponding author: Paul L Hofman, Liggins Institute, University of Auckland, Private Bag 92019, Auckland, New Zealand. (+64 9) 923 6453. [email protected] URL: http://www.nzma.org.nz/journal/read-the-journal/all-issues/2010-2019/ 2015/vol-128-no-1416/6564
REFERENCES: 1. Assaf SA, Randolph LM, Randolph L (2013) Chime- 7. Hughes IA, Houk C, Ahmed Benirschke K, Wu S, rism in monochorionic SF, Lee PA, LWPES Consen- Samadi R Chmait RH dizygotic twins: case study sus Group ESPE Consensus (2010) Discordant blood and review. Am J Med Group (2006) Consensus chimerism in dizygotic Genet A 161A:1817-1824. statement on management monochorionic laser-treat- 5. Crofton PM, Evans AEM, of intersex disorders. Arch ed twin-twin transfusion Groome NP, Taylor MRH, Dis Child 91:554-563. syndrome. Obstet Gynecol Holland CV Kelnar CJH 8. Johannsen TH, Lundsteen 116 Suppl 2:483-485. (2002) Dimeric inhibins in C, Visfeldt J, Schwartz M, 2. Barbaro M, Wedell A girls from birth to adult- Petersen BL, Byskov AG Nordenström A (2011) hood: relationship with age, Muller J (2003) Erroneous Disorders of sex devel- pubertal stage, FSH and genetic sex determina- opment. Semin Fetal oestradiol. Clin Endocri- tion of a newborn twin Neonatal Med 16:119-127. nol (Oxf) 56:223-230. girl due to chimerism 3. Bourthoumieu S, Esclaire 6. Hawcutt D, Hammond B, caused by foetal blood F Yardin C (2006) Chime- Sibbring J, Gokhale D, Ellis transfusion. A case report. rism in twins: caution is I, Bricker L Subhedar N Horm Res 60:148-151. needed in interpretation (2011) Twin-twin confu- 9. van Dijk BA, Boomsma DI of karyotypes. Am J Med sion syndrome: blood de Man AJ (1996) Blood Genet A 140:533-535. chimerism in opposite sex group chimerism in human 4. Chen K, Chmait RH, dizygotic twins. J Obstet multiple births is not rare. Vanderbilt D, Wu S Gynaecol 31:446-448. Am J Med Genet 61:264-268.
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 65 www.nzma.org.nz/journal VIEWPOINT
Domperidone safety: a mini-review of the science of QT prolongation and clinical implications of recent global regulatory recommendations Pamela J Buffery, R. Matthew Strother
ABSTRACT AIMS: In New Zealand, domperidone is approved for gastrointestinal motility and nausea and vomiting. The European Medicines Agency (EMA) recently concluded that domperidone poses a significant risk of sudden cardiac death (SCD) and has restricted use in Europe. This paper reviews the risk of QT prolonga- tion and cardiac adverse effects with domperidone and provide information to allow prescribers to make informed decisions on usage. METHODS: A search of two bibliographic databases, the European Medicines Agency (EMA) website, Micro- medex, Lexicomp and reference texts was undertaken for domperidone related reports of QT prolongation, cardiac arrhythmias and/or SCD. The New Zealand Centre for Adverse Drugs Reaction Monitoring was also contacted for cardiac adverse event reports with domperidone. RESULTS: Over 30 published papers, EMA documents and other information sources were collated, including two studies that met thorough QT study (TQT) criteria (ICH-E14). The first TQT1 was negative while the second was marginally positive. Reports of QT prolongation, ventricular arrhythmias and SCD were located (predominantly high/very high–dose IV domperidone). With oral domperidone, a Dutch case– controlled study reported an adjusted odds ratio of SCD of 11.4 (95% CI 1.99-65.2), based on only three patients out of 1,366 cases of SCD. A second nested case–controlled study calculated an odds ratio of ventricular arrhythmia or SCD of 1.59 (1.28–1.98) vs. placebo. DISCUSSION: Based on the results of the two TQT (the regulatory agency gold standard for assessment of QT prolongation) domperidone does not appear to be strongly associated with QT prolongation at oral doses of 20 mg QID in healthy volunteers. Further, there are limited case reports supporting an association with cardiac dysfunction, and the frequently cited case-control studies have significant flaws. While there remains an ill-defined risk at higher systemic concentrations, especially in patients with a higher baseline risk of QT prolongation, our review does not support the view that domperidone presents intolerable risk.
groups across New Zealand because it is Introduction believed to be both efficacious and have Domperidone is licensed in New Zealand a relatively limited side-effect profile, as a treatment for nausea, vomiting, particularly compared to other dopamine and dyspepsia.1 In practice, it is used as receptor blockers (eg, metoclopramide), a pro-kinetic, an adjunct in Parkinson’s insofar as incidence of extrapyramidal side disease, treatment of chemotherapy effects.2 However, the recommendations induced nausea and vomiting, and off-label recently published by the European as a galactagogue. Domperidone is a Medicines Agency (EMA) and a subsequent preferred medication by several practice review by MedSafe’s Medicines Adverse
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 66 www.nzma.org.nz/journal VIEWPOINT
Table 1: Medicine Adverse Reaction Committee (MARC) Recommendations June 2014
The Committee recommended that Medsafe request the sponsors of domperidone to update the 1 indications, contraindications and warnings and precautions sections of the data sheets.
The Committee recommended that Medsafe request the sponsors of domperidone to update the 2 data sheets with a reduced recommended daily dose for use in children.
The Committee recommended that Medsafe communicates the results of this review to healthcare 3 professionals by including an article on domperidone in a future edition of Prescriber Update.
The Committee recommended that Medsafe communicates the results of this review directly to 4 nurses and midwives through the New Zealand Nurses Organisation (NZNO) and New Zealand College of Midwives.
Reactions Committee (summarised in treatment of nausea and vomiting; and use Table 1), may change the way New Zealand in high risk groups (hepatic dysfunction, the practitioners use this drug. In this paper elderly (≥60 years), those with pre-existing we explore the background for concern QT prolongation, and in those on other QT by global drug regulatory agencies and prolonging agents) should be curtailed.3 critically examine the strength of the These regulatory restrictions are consistent, evidence behind their recommendations. though less onerous, with the American Domperidone is a peripheral dopamine approach, where domperidone is only D2-receptor antagonist with prokinetic and available via special application to the Food antiemetic properties that has been available and Drug Administration (FDA) as an inves- since 1978.3 Studies have explored a variety tigational agent.5 of indications: nausea and vomiting (both related and unrelated to chemotherapy); Sudden cardiac death dyspepsia; bowel motility disorders, and control of the gastrointestinal side effects of and QT prolongation anti-Parkinson medications.4 Initial reports The concern driving drug regulatory of QT prolongation and cardiac events agencies is an association between domper- started to appear in the 1980s and related to idone and SCD. While the exact aetiology administration of high-dose intravenous (IV) of SCD frequently cannot be determined, domperidone. As a direct consequence of regulators believe domperidone leads to these case reports domperidone was never progressive prolongation of the QT interval. approved in the US and the IV formulation The QT interval is considered a biomarker was withdrawn globally. However, oral for SCD because it has been observed with domperidone remains available in many prolongation, the cardiac depolarisation/ countries, including New Zealand. repolarisation cycle becomes more chaotic, In 2013 the EMA, at the request of the increasing the likelihood of a depolarisation Belgian drug regulatory agency, undertook occurring concurrently with repolari- a review of the clinical use and safety of sation, leading to ventricular dysrhythmia domperidone. The report from the Phar- Torsades de Pointes (TdP), which has a high 6 macovigilance Risk Assessment Committee, risk of SCD. Indeed, families with genetic released in March 2014, found that mutations of cardiac ion channels reveal domperidone posed a significant risk of QT that individuals with persistent QT prolon- prolongation, cardiac dysrhythmias, and gation have a significantly higher risk of sudden cardiac death (SCD). Citing this risk, TdP and SCD, as can electrolyte abnor- and concerns over efficacy data, the EMA malities, other medical conditions, and 7 made the following recommendations: importantly, medications. maximum doses of domperidone in adults The molecular mechanism of drug-in- and children should be reduced to 30 mg/ duced QT prolongation has not been fully day and 0.75 mg/kg/day, respectively; use elucidated, but is linked to the potassium should be restricted to less than 1 week; efflux channel, human Ether-a-go-go-Related domperidone should only be used for Gene (hERG).8,9 A drug can bind to the hERG
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 67 www.nzma.org.nz/journal VIEWPOINT
channel effectively closing it, thus delaying searched for any relevant data regarding cardiac repolarisation and lead to QT prolon- domperidone and QT prolongation, cardiac gation. Unfortunately, the transmembrane arrhythmias and/or sudden cardiac portion of hERG is promiscuous, binding to death. We also browsed and searched the many drugs, and thus hERG-mediated QT EMA website for documents relating to prolongation has been associated with many their recent recommendations regarding drugs.10 Non-trivial risk for QT prolongation domperidone. Lastly, we contacted the New exists for many drugs and potential drugs Zealand Centre for Adverse Drugs Reaction and significant investment has been put into Monitoring for the numbers of reports the pre-clinical and clinical quantification received by the centre relating to the use of this risk. A number of pre-clinical in vitro of domperidone and reports of QT prolon- and in vivo models have been developed, gation, cardiac arrhythmias and SCD. and extensive reviews of advances in this area can be found in the literature.11,12 For Domperidone and purposes of this review, however, it is signif- icant to note that while regulatory guidance SCD: initial reports has been non-directive on species selection, Concerns over the cardiac effects of generally dog is considered the in vivo domperidone arose in the 1980s when five model of choice in relation to humans for QT cases of cardiac arrest were reported with prolongation assessment.13 use of high–dose IV domperidone. The initial Clinically, the gold standard for deter- report was of a single patient in a small dose- mining a drug’s risk for QT prolongation escalation trial for chemotherapy-induced and SCD is unambiguously the ‘thorough QT’ nausea and vomiting who developed fatal study (TQT).14 The International Committee ventricular fibrillation following a 200 mg on Harmonisation (ICH) E14 Guidance on IV bolus of domperidone.16 In spite of this Clinical QT Studies has been adopted by all death, the study reported on a further three major drug regulatory bodies. This guidance patients administered between 0.94 mg/ recommends a basic study design including kg and 13.4 mg/kg given divided between a positive control (ie, a fluoroquinolone) an initial IV bolus followed by a 13-hour that will increase the QTc by about 5ms in infusion. The authors reported tolerance and most subjects, sufficient dose levels to allow good efficacy at 1.2 mg/kg IV bolus followed assessment of a dose-response curve to QTc by 1.6 mg/kg IV infusion. A subsequent prolongation, and defines the negative TQT series of four case reports occurred at as a study in which the largest baseline-sub- more moderate doses of 50 mg (one over tracted, time-matched, mean difference 2 hours and repeated after chemotherapy, between drug and placebo is ≤ 5 ms with an the other a single dose over 15 minutes) upper bound of the 95% confidence interval and 20 mg (both as IV bolus).17 However, of ≤ 10 ms.15 While there remains signif- it was noted that three of these patients icant debate around aspects of the TQT—the had hypokalaemia, a known risk factor for impact of sex, cross-over versus parallel ventricular dysthymias, all were successfully design, and methodologies/algorithms for QT resuscitated, and one of the patients went interval determination and rate correction on to receive further IV domperidone with (determining the QTc)—the negative TQT no sequelae. However, based on these is generally accepted that the drug imparts concerns the regulatory authorities removed minimal risk for TdP and SCD. IV domperidone from global markets, and formal studies of cardiac risk with Methods domperidone were initiated. A literature search of Embase (1947 to present) and Medline (1946 to present) Oral domperidone was conducted using various Medical in the thorough Subject Headings (MeSH) terms that included: domperidone; QT prolongation; QT study heart arrhythmia; QT prolongation.mp; Two TQT studies have been performed and arrhythmias-cardiac. In addition, to assess the risk of oral domperidone. several databases and reference texts were The first, cited in the EMA report, was a
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 68 www.nzma.org.nz/journal VIEWPOINT
well-designed four-way crossover study prescription for domperidone at the time with placebo and moxifloxacin as an active of their event.19 Matched against 20 or 40 control in 44 healthy volunteers.3 Two controls, depending on whether the event dosage regimens of domperidone were was fatal the odds ratio (OR) for VA/SCD compared, 10 mg four times a day (40 mg/24 due to domperidone was just under four. hours) and 20 mg four times a day (80 mg/24 Further subgroup analysis of the three hours), with drug concentration monitoring cases with domperidone doses >30 mg/day of both domperidone and moxifloxacin had an adjusted OR (aOR) of 11. conducted during the study. Three methods The second study20 matched 140 cases were used to determine QTc. No clinically with 560 controls not on QTc prolonging relevant dose-QTc response and exposure drugs and found an aOR of 2.1 (1.2–3.5) QTc response effects were observed in this for patients on any medication associated study. This was a negative study according to with QTc prolongation, and 4.7 for domper- the ICH-E14 criteria with oral domperidone idone specifically. This compares to aORs at single of doses 10 to 20 mg, or multiple of 3.87 (1.6-9.2), 2.6 (1.4-6.4), 2.0 (0.5-8.1), doses of 40 to 80 mg in 24 hours. and 1.4 (0.2-8.6), for haloperidol, cotrimox- The second study, not cited by the azole, amitriptyline, and clarithromycin, EMA, was a pharmacokinetic and TQT respectively. Taking more than one drug double-blind crossover study in 24 associated with QTc prolongation had an healthy volunteers.18 Study arms were: aOR of 4.8 (1.6-14), and specifically taking domperidone (10 mg four times a day); drugs that inhibited metabolism of a ketoconazole (200 mg twice daily); and a concurrently prescribed QTc prolonging combination of domperidone and keto- drug had an aOR of 4 (1.2-13). conazole with placebo. Monitoring and A Canadian nested case-controlled study21 determination of QTc was similar to that of using a healthcare database of just under the first study. QT prolongation greater than 1 million residents compared the risk of 470 ms was not observed with any single proton pump inhibitors (PPIs) and domper- drug or combination. The highest mean idone for SCD. The 83,212 individuals increase in QTc from baseline compared prescribed domperidone, a PPI, or both with placebo was the combination at 15.9 were matched with up to four controls, and ms ([95% CI, 12.47 to 19.33], p<0.001); 1,608 cases of VA or SCD were identified. Of ketoconazole alone was 9.24 ms ([5.85 to these, 169 (10% of those having VA/SCD) had 12.63] p<0.001); and domperidone was 4.2 active domperidone prescriptions at the ms ([0.77 to 7.63], p=0.017). Therefore, this time of their event. Compared to current study is borderline positive for domper- PPI use domperidone had an OR of 1.44 idone monotherapy (maximum mean (1.12–1.86), and compared to placebo 1.59 increase in QTc compared with placebo (1.28–1.98). was > 5 ms at some time points but never exceeded 12 ms). It should be noted, The New Zealand Pharmacovigilance however, that this increase is in the same Centre has received a total of 32 spontaneous range as the accepted positive control, an reports of any type of adverse effect asso- azole antifungal, where the risk is felt to be ciated with the use of domperidone submitted minimal enough that these medications are by various health professionals around New routinely prescribed not just in illness, but Zealand. Only one of these reports related in healthy volunteers in TQT studies. to QT prolongation, from which the patient was reported to recover from (no further Case-controlled details known). There were no other reports for QT prolongation and none for SCD, TdP, studies or arrhythmia of any kind. From a global Three case-control studies have perspective, as of June 2014, the World Health reported an increased risk of ventricular Organization adverse drug event database arrhythmias (VA) or SCD associated with had a total of 67 reports (six fatal) of domper- domperidone use, summarised in Table 2. idone associated with QT prolongation, 34 of Van Noord et al. identified 1,336 cases of VA TdP (6 fatal), 21 of arrhythmia (one fatal) and or SCD of which 10 (0.75%) had a current 41 of cardiac arrest (29 fatal).22
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 69 www.nzma.org.nz/journal VIEWPOINT
Table 2: Case control studies of domperidone and VA or SCD Study Control Number of Cases Population Population on Domperidone OR (95% CI) Electronic primary care 10 with domperidone record with 1 million Patients without 3.72 (1.7.2–8.08) 3 with domperidone > patients VA or SCD 11.4 (1.99–65.2) 30mg/24 h [van Noord C, et al, 19]
Patients with Single institution data- cardiac arrest or 140 cases on a QTc prolong- base of cardiac arrest SCD not on any ing drug 4.7 (1.4–16) and SCD over 8 years QTc prolonging 7 cases on domperidone [De Bruin ML, et al, 20] medications
1.44 (1.12– 1.86) compared to PPI Electronic database of Patients without 1.59 (1.28– 1.98) approximately 1 million VA or SCD with or 69 cases on domperidone compared to no patients without PPI exposure to either [Johannes CB, et al, 21] domperidone or a PPI
First, the case reports that prompted Discussion the initial concern over SCD were asso- Domperidone has been available for ciated with high–dose IV domperidone. prescribing in New Zealand since 1984, Indeed, given that domperidone has oral and in that time CARM has had no reported bioavailability of 13–17% the reported cases of VA or SCD associated with its use. cases would have occurred at oral doses However, domperidone has received the in excess of 1000 mg in the initial case, attention of global drug regulatory agencies or at bolus doses of 150 mg to 380 mg.4 because of association with these very Second, the two presented TQT studies serious risks. give a mixed picture—one is negative, Before summarising the clinical liter- with dosing up to 80 mg/24 hours and the ature, it is worth noting that a significant second is marginally positive, with dosing component of the EMA recommendation up to 40 mg/ 24 hours. It is interesting was based on pre-clinical data. This is that the marginally positive study was not remarkable because the E14 guidance cited in the EMA literature, but it is worth preferences clinical data above preclinical. noting that the QTc prolongation asso- In fact, E14 specifically states that only in ciated with domperidone was similar to the case of an ambiguous TQT and signif- that of routinely prescribed fluoroquino- icant clinical concern should animal data lones and many other QT prolonging drugs be used to guide clinical decisions. It is not associated with extensive regulatory therefore discordant that the EMA cites restrictions on their use. The case-control a negative TQT, but based a significant studies using retrospective databases do portion of their recommendations on consistently identify a risk of VA/SCD with animal model data performed in rodents, a domperidone—although the extent of that non-preferred model for human hERG.3 So, risk varies wildly with the more extreme this review focuses on the available clinical estimates being driven by a very small information to help the clinician interpret number of cases. Finally, it is important to these data. note that in New Zealand there is one report The clinical literature defining this risk potentially associating domperidone with with oral domperidone is comprised of a QT prolongation (non-fatal); globally, only negative TQT, a marginally positive TQT 67 spontaneous reports (42 fatalities) are and three case control studies. It is worth on record associating domperidone with remarking on some aspects of these studies cardiac events from 1984 to 201422—bearing to address the concerns raised by the EMA in mind that these data are not extensively and subsequently New Zealand Medsafe. validated and prone to over-association.
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 70 www.nzma.org.nz/journal VIEWPOINT
In summary, this literature is mixed, and Alternatives to domperidone as an therefore, as with the many approved and antiemetic or as a prokinetic, the two routinely prescribed medication associated indications licensed in New Zealand, are with QT prolongation (the reader is referred presented in Table 3. Given that both to CredibleMeds, www.crediblemeds.org), nausea/vomiting and motility disorders prescribing clinicians should be aware of have presently available alternatives, this potential risk and work to mitigate this prescribing clinicians must weigh the risks or pursue alternatives. Limiting risk means and benefits of prescribing domperidone knowing the other clinical contributors and select this as the drug of choice. A regu- to QTc prolongation, including: baseline latory restriction on maximum dosage will QTc prolongation (genetic, structural, or mandate prescribers choose either a less drug-induced); electrolyte imbalances effective dose or an alternative, regardless (hypo and hyper-kalemia); concurrent of whether they feel the risks and benefits prescription with other QTc prolonging are in favour of domperidone. We would drugs; or concurrent prescription with argue given the relative weakness of data drugs inhibiting the metabolism of domper- driving the decisions of the EMA and the idone (cytochrome p450 3A).8,23,24 Indeed, FDA, that New Zealand prescribers should it was notable that the second TQT study acknowledge a possible risk, but have the presented observed significantly greater ability to weigh this risk both in the context domperidone exposure (2.9 and 3.6 fold of the clinical decision and in light of alter- higher Cmax and AUC, respectively) as natives, rather than be forced into avoiding a result of the CYP3A inhibition by keto- domperidone. conazole. If however, the clinician is uncomfortable with this risk, they should look to alternatives.
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 71 www.nzma.org.nz/journal ARTICLE
Table 3: Alternative agents to domperidone Drug Selected common ADRs (%) Indication Selected serious ADRs (%) (QT prolongation risk) Aprepitant neutropaenia (up to 9%) alopecia (24%) hiccup (≤11%) bradycardia (4%) fatigue (≤22%) diarrhoea (≤10%) antiemetic sinus tachycardia (cr) somnolence (≤20%) anorexia (≤10%) (NO) SJS (cr) muscle weakness (≤18%) hypotension (6%) TEN (cr) headache (≤16%) puritus (8%) neutropaenic sepsis (cr) nausea (≤13%) fever (≤6%) constipation (≤12%) Cyclizine raised ventricular filling pressures drowsiness (>10%) nausea (≤10%) (ind) xerostomia (>10%) urinary retention (≤10%) antiemetic palpitations (ind) headache (≤10% diplopia (≤10%) (NO) tachycardia (ind) dermatitis (≤10%) cholestatic jaundice (cr) agranulocytosis (cr) Dexamethasone bradyarrhythmia (ind) hypertension (ind) hypokalaemia (ind) cardiac dysrhythmia (cr) impaired wound healing (ind) osteoporosis antiemetic cardiomyopathy (ind) acne (ind) (NO) thromboembolic disorder (ind) skin thinning (ind) pancreatitis (ind) hyperglycaemia (ind) psychiatric symptoms (inn ind convulsions (ind) Cushing’s syndrome gastrointestinal ulcer (ind) immunosuppression (ind) anaphylaxis (ind) aseptic necrosis of bone pseudotumour cerebri (cr) raised intracranial pressure (cr) Droperidol/Haloperidol cardiac arrest (ind) hypertension (ind) tachycardia (ind) hypotension (ind) antiemetic TdP (ind) hyperprolactinaemia (ind) (YES) ventricular tachycardia (ind) somnolence (ind) extrapyramidal symptoms (ind) hallucinations (ind) hyperactivity (ind) NMD (cr) psychiatric symptoms (ind) haematological effects (ind) Erythromycin TdP (ind) diarrhoea Sudden cardiac death (ind) anoreaxia prokinetic ventricular arrhythmia (ind) nausea (YES) convulsions (ind) vomiting erythema multiforme (rare) SJS/TEN (rare) C.difficle colitis haematological effects (ind) hepatotoxicity (ind) anaphylaxis (rare) dystonia (rare) interstitial nephritis (cr) hearing loss 9cr) Lorazepam extrapyramidal symptoms (<1%) sedation (≤16%) hepatotoxicity (<1%) dizziness (≤7%) antiemetic cerebral oedema (<1%) (NO) cardiac arrest (<1%) convulsions (<1%) psychiatric symptoms (<1%) GI haemorrhage (<1%) haematological effects (<1%) NMS (<1%) paralysis (<1%) pericardial effusion (<1%) Metoclopramide dystonia (≤25%, dose dependent) drowsiness (≤70%, dose dependent) psychiatric symptoms (cr) fatigue (≤10%) antiemetic/prokinetic cardiac arrest (cr) nausea (≤6%) (NO) cardiac dysrhythmia (cr) headache (≤5%) malignant hypertension (cr) dizziness (4%) supraventricular tachycardia (ind) extrapyramidal symptoms (ind) haematological effects (ind) hepatotoxicity (rare)
Ondansetron bradycardia (<1%) headache (≤27%) cardiac arrhythmia (<1%) fatigue (≤13%) antiemetic cardiac arrest (<1%) constipation (≤13%) (YES) extrapyramidal symptoms (<1%) drowsiness (≤8%) hepatic failure (<1%) dizziness (≤7%) Key: hepatic necrosis (cr) diarrhoea (≤16%) NMS (<1%) fever (≤8%) ADR adverse drug reaction TEN (<1%) xerostomia (5%) cr case reports TdP (cr) LFT elevation (17%) anaphylaxis (cr) ind incidence not defined Prochlorperazine cardiac arrest (ind) dizziness (ind) SJS Stevens-Johnson Syndrome sudden cardiac death (ind) somnolence (ind) antiemetic TEN toxic epidermal necrolysis NMS (ind) (YES) haematological effects (ind) NMS neuroleptic malignant syndrome hepatotoxicity (ind) GI gastrointestinal extrapyramidal symptoms (ind) Torsades de Pointes tardive dyskinesia (ind) TdP convulsions (jnd) LFT liver function tests psychiatric symptoms (ind)
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 72 www.nzma.org.nz/journal VIEWPOINT
Competing interests: Nil Author information: Pamela J Buffery: Chief Drug Information Pharmacist, Department of Clinical Pharmacology, Canterbury District Health Board, Christchurch; R. Matthew Strother: Oncologist, Department of Oncology, Canterbury District Health Board, Christchurch Corresponding author: Pamela Buffery, Department of Clinical Pharmacology, Christchurch Hospital, Private Bag 4710, Christchurch 8140, New Zealand [email protected] (+64 3) 364 0900 URL: http://www.nzma.org.nz/journal/read-the-journal/all-issues/2010-2019/ 2015/vol-128-no-1416/6565
REFERENCES: 1. New Zealand Formulary. 8. van Noord C, Eijgelsheim prolongation) by human Domperidone. New M, Stricker BHC. Drug- and pharmaceuticals S7B. 2005 Zealand Formulary Online non-drug-associated QT (www.physiostim.com/ (www.nzf.org.nz) interval prolongation: download/ICHS7B.pdf) 2. Barone JA. Domperidone: QT interval prolongation. 14. FDA. Guidance for Indus- a peripherally acting Br J Clin Pharmacol. try: E14 Clinical Evaluation dopamine2-receptor 2010;70:16-23 of QT/QTc Interval Prolon- antagonist. Ann Pharma- 9. Vandenberg JI, Perry MD, gation and Proarrythmic cother. 1999;33:429–440 Perrin MJ, Mann SA, et Potential for Non-Antiar- 3. EMA. Pharmacovigi- al. hERG K+ channels: rhythmic Drugs. 2012 lance Risk Assessment structure, function, and 15. FDA. Guidance for Indus- Committee. Domperidone clinical significance. Physiol try: E14 Clinical Evaluation Assessment Report. Rev. 2012;92:1393-478 of QT/QTc Interval Prolon- EMA/ 152501/ 2014. 10. CredibleMeds. Combined gation and Proarrythmic 4. Brogden RN, Carmine AA, List of All QTdrugs and the Potential for Non-Antiar- Heel RC, et al. Domper- List of Drugs to Avoid for rhythmic Drugs. 2005 idone. A review of its Patients with Congenital 16. Letellier N, Coulomb pharmacological activity, Long QT Syndrome. 2014 F, Lebec C, Brumet JM. pharmacokinetics and (www.crediblemeds.org) Sudden death in cancer therapeutic efficacy in the 11. Valentin JP, Pollard C, patient on high-dose symptomatic treatment Lainée P, Hammond T. domperidone. Lancet. of chronic dyspepsia Value of non-clinical 1982;319:1019 and as an antiemetic. cardiac repolarization 17. Roussak JB, Carey P, Drugs 1982;24:360–400 assays in supporting the Parry H. Cardiac arrest 5. FDA Talk Paper: FDA discovery and develop- after treatment with Warns Against Women ment of safer medicines: intravenous domperidone. Using Unapproved Drug, Non-clinical cardiac BMJ. 1982;289:1579 repolarization assays. Br J Domperidone, to Increase 18. Boyce MJ, Baisley KJ, Pharmacol. 2010;159:25-33 Milk Production. 2004 Warrington SJ. Phar- 6. Ritter JM. Cardiac safety, 12. Wallis RM. Integrated risk macokinetic interaction drug-induced QT prolon- assessment and predic- between domperidone gation and torsade de tive value to humans of and ketoconazole leads pointes (TdP) Br J Clin non-clinical repolarization to QT prolongation in Pharmacol. 2012;73:331-4 assays. Br J Pharmacol. healthy colunteers: a 2010;159:115-21 7. Sauer AJ, Newton-Cheh randomized, placebo-con- C. Clinical and Genetic 13. ICH Working Group. trolled, double-blind, Determinants of Torsade The non-clinical eval- crossover study. Br J Clin de Pointes Risk. Circula- uation of the potential Pharmacol. 2012;73:411-21 tion. 2012;125:1684–94 for delayed ventricular 19. van Noord C, Dieleman repolarization (QT interval
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 73 www.nzma.org.nz/journal VIEWPOINT
JP, van Herpen G, et 21. Johannes CB, Varas-Lorenzo for Adverse Reactions al. Domperidone and C, McQuay LJ, et al. Risk Monitoring (16/06/2014) ventricular arrhythmia of serious ventricular 23. Heist EK, Ruskin JN. or sudden cardiac death. arrhythmia and sudden Drug-Induced Arrhyth- Drug Safety. 2010;33:881-8 cardiac death in a cohort mia. Circulation. 20. De Bruin ML, Langendijik of users of domperidone: a 2010;122:1426–35 nested case-control study. PMJ, Koopmans RP, et al. 24. Yoshizato T, Kotegawa Pharmacoepidemiology and In-hospital cardiac arrest T, Imai H, et al. Itracon- Drug Safety. 2010;19:881-8 is associated with the use azole and domperidone: of non-antiarrhythmic 22. Personal Communication a placebo-controlled QTc-prolonging drugs. with the Director New drug interaction study. Br J Clin Pharmacol. Zealand Pharmacovigilance Eur J Clin Pharmacol. 2006;63:16-23 Centre. New Zealand Centre 2012;68:1287–94
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 74 www.nzma.org.nz/journal CLINICAL CORRESPONDENCE
Not a simple back pain Jen-Li Looi, Ruvin Gabriel
61-year-old man, with previous suggestive of an infected endovascular endovascular stenting from stent with abscess formation. Surgery was A distal aortic arch to below the not an option, due to multiple medical diaphragm for subacute type B aortic comorbidities and abscess was unable to dissection in 2010, presented with acute be drained percutaneously. He continued coronary syndrome and subsequently to deteriorate and eventually passed developed intravenous line infection on away. This report illustrates the diagnostic day 5 of admission. He also described a capability of CT in providing incremental dull ache in the mid scapular region at information about the complications the same time while he developed the associated with prosthetic aortic graft intravenous line infection. He had ongoing infection for guiding treatment. Staphylococcus aureus bacteraemia despite intravenous antibiotics. Transoesophageal Learning points echocardiogram showed no vegetations. • Prosthetic aortic graft infections Contrast CT chest (Panel B and C) represent a major diagnostic and demonstrated a soft tissue rind (yellow therapeutic challenge associated with arrows) around the descending thoracic considerable morbidity and mortality. aorta stent (red arrow) from the level of • Computed tomography (CT) is the 4th thoracic vertebra down to the aortic the imaging modality of choice in hiatus, which was new in comparison to the investigation of patients with the previous scan performed 12 months suspected prosthetic aortic graft earlier (Figure A). The appearances were infection
Figure: A: There was no soft tissue thickening around the thoracic stent on the previous scan performed 12 months earlier.
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 75 www.nzma.org.nz/journal CLINICAL CORRESPONDENCE
Figure B: Contrast CT chest demonstrated a soft tissue rind (yellow arrows) around the descending thoracic aorta stent (red arrow) Figure C: Again, there was soft tissue thickening (yellow arrows) around the descending thoracic aorta stent noted on contrast CT chest on the sagittal plane.
Competing interests: Nil Author information: Jen-Li Looi, Department of Cardiology, Middlemore Hospital, Private Bag 933111, Otahuhu, Auckland, New Zealand; Ruvin Gabriel, Department of Cardiology, Middlemore Hospital, Private Bag 933111, Otahuhu, Auckland, New Zealand. Corresponding author: Jen-Li Looi, Department of Cardiology, Middlemore Hospital, Private Bag 93311, Otahuhu, Auckland, New Zealand. (+64 9) 276 0061. [email protected] URL: http://www.nzma.org.nz/journal/read-the-journal/all-issues/2010-2019/ 2015/vol-128-no-1416/6566
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 76 www.nzma.org.nz/journal CLINICAL CORRESPONDENCE
Recalcitrant peripheral spondyloarthritis treated with radiotherapy Nur Azri Bin Haji Mohd Yasin, Shaun Anthony Costello
28-year-old-woman was referred With the first course, she was symptom-free for consideration of radiotherapy to for a period of 8 months. A second course was A her right knee, affected by persistent given after relapse and she was again symp- enthesitis secondary to spondyloarthritis. tom-free for several months, enabling her to She presented with recurrent episodes be mobile for her wedding day. of severe debilitating right knee pain At the end of the second course, her only on weight-bearing and movement, side effect was mild erythema of the irra- necessitating her to wear a leg brace and diated skin. crutches to walk. She had been admitted multiple times for management of her pain and these episodes recur every 2–3 months. Discussion Spondyloarthritis is a relatively common The pain was located on the lateral aspect condition with prevalence of 0.4 to 1.9%.1 In of her right knee. The area was swollen, this case report, radiotherapy was utilised exquisitely tender, with a degree of dyses- to palliate the patient’s pain to good effect. thesia. She also developed significant Radiotherapy was a mainstay treatment for valgus angulation at the right knee with spondyloarthritis in the 1930s to 1950s but consequent valgus angulation at the ankle, it was soon abandoned due to its associ- forefoot pronation and clawing of the toes, ation with secondary leukaemia.2 secondary to pain. Increased leukaemia risk is a late compli- During flare-up, the CRP was raised and cation of spinal irradiation with significant on MRI, enthesitis of the biceps femoris volumes of haematopoietically active bone with associated reactive synovitis was seen. marrow receiving large mean radiation The enthesitis was also evident on nuclear doses.3 Theoretically, localised radiotherapy medicine scan. to peripheral spondyloarthritis should She was managed with multiple ther- not have the same risk, as the volume of apies, which include local steroid injection, irradiated bone marrow is minimal and methotrexate, sulphasalazine, mesalazine, the bone marrow in the distal long bones leflunomide, IV methylprednisolone, phys- in an adult is not usually haematopoieti- iotherapy and acupuncture. Unfortunately, cally active. Also, it should be noted that the none of this was consistently helpful in radiotherapy given in the period when radi- improving her symptoms. ation was routinely used was orthovoltage She had two courses of radiotherapy to the radiation, which is absorbed unequally by same area; 15Gy in 10 fractions over 2 weeks. bone and soft tissues.2 Modern megavoltage
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 77 www.nzma.org.nz/journal CLINICAL CORRESPONDENCE
Figure 1: Three-dimensional CT images of plan
radiotherapy has the advantage of being In the literature, evidence for localised absorbed equally by bone and soft tissues radiotherapy for peripheral spondyloar- and advancements in technology especially thritis is old, but was shown to be effective with the implementation of CT planning, in relieving pain.3,5 The dose reported in radiation delivery allows greater precision those papers were 10–20Gy course in 1–2Gy with minimal doses to normal tissues. We fractions.3,5 Our case is a recent experience believe the main secondary malignancy and has showed that localised radiotherapy risk in localised radiotherapy is basal cell is effective for recalcitrant peripheral carcinoma, with an absolute lifetime risk of spondyloarthritis and in selected cases, we approximately 1 in 105 for 1cm2 per Gy.4 believe it should be considered as a valid option for palliation.
Competing interests: Nil Author information: Nur Azri Bin Haji Mohd Yasin, Radiation Oncology Department, Dunedin Public Hospital, Southern DHB, Dunedin; Shaun Anthony Costello, Radiation Oncology Department, Dunedin Public Hospital, Southern DHB, Dunedin. Corresponding author: Shaun Anthony Costello, Radiation Oncology Department, Dunedin Hospital, Private Bag 1921, Dunedin 9054, New Zealand. (+64 3) 474 0999 [email protected] URL: http://www.nzma.org.nz/journal/read-the-journal/all-issues/2010-2019/ 2015/vol-128-no-1416/6567
REFERENCES: 1. Bakland G, Nossent 3. Grill V, Smith M, Ahern diseases. Strahlenther HC. Epidemiology of M, Littlejohn G. Local Onkol. 2006;182(8):431-6. spondyloarthritis: a radiotherapy for pedal 5. Mantell BS. Radiotherapy review. Curr Rheumatol manifestations of for painful heel syndrome. Rep. 2013;15(9):351 HLA-B27-related arthrop- Br Med J. 1978;2(6130):90-1. 2. Muscoplat CC, Caperton EM, athy. Br J Rheumatol. Dusenbery KE. Radiation 1988;27(5):390-2. therapy for inflammatory 4. Trott KR, Kamprad F. Esti- arthritis. Int J Radiat Oncol mation of cancer risks from Biol Phys. 2004;60(2):688-9. radiotherapy of benign
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 78 www.nzma.org.nz/journal LETTER
Direct access GP referral for ETT functions as a virtual clinic Jessica H Greaves, Joan D Leighton, John G Lainchbury, Paul G Bridgman
he Canterbury District Health Board Four DAETT referrals were rejected. Two offers a service allowing GPs to of these were advised to have further GP Trefer a patient directly for a Direct investigations. One did not have enough Access Exercise Tolerance Test (DAETT). information on the referral and 1 was The service is for patients with chest pain requested as a screening tool in an asymp- who have been assessed as having a low tomatic patient and therefore declined as probability of angina. Criteria for this were not meeting criteria. locally developed in a primary and second- Thirty-nine DAETTs were performed. ary care collaboration as part of Canterbury Thirty-three (84%) of these were negative. HealthPathways. All referrals are triaged by Three (8%) tests were reported as equivocal the Cardiology Liaison GP, who can either and 3 tests were positive. Of the positive accept the referral, upgrade to an outpatient tests, 1 patient was admitted and went on Cardiology appointment if the patient is to have CABG, 1 had a coronary angiogram felt to be high risk for angina, or reject the which showed severe stenosis in modest referral if criteria are not met. sized branch vessels, and 1 underwent We prospectively followed 52 consecutive CT coronary angiogram, which showed patients referred for DAETT in October minimal coronary artery disease. Of the 2014. The mean age was 60 years (range equivocal tests, 1 has been referred by GP 38–84), 25 were male, 4 were smokers, and for spirometry, 1 has ongoing symptoms 3 had diabetes. In 15 of the referrals the GP that the GP feels are not cardiac, and 1 has had a taken TnI pre-referral. All of these resolved symptoms. were negative. All 52 study patients were followed up Of the 52 referrals, 43 (83%) were accepted with a phone call to their GP practice 20 for a DAETT. Thirty-nine of these went weeks following the initial referral. In this ahead; 1 patient opted for a private test, 1 time, the majority (58%) of the patients with was admitted to Christchurch Public Hospital a negative DAETT had been back to their GP. in the interim for an overall decline in his In only 2 cases was there ongoing concern health and 2 patients cancelled their tests as about the symptoms that had led to the they reported the problem had resolved. initial DAETT referral. One was now diag- Five DAETT referrals were upgraded nosed as dyspepsia and 1 was referred back to Cardiology outpatient appointments. to Cardiology as the GP had been advised to Three of these had further testing do so if ongoing symptoms as the negative showing CAD. Of the other 2, 1 patient DAETT was submaximal. had a negative dobutamine stress echo These results show that the service is and was discharged from follow up, and 1 working well for patients and GPs. The cancelled their appointment as symptoms patients with negative test results are had resolved. almost always well reassured with a low
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 79 www.nzma.org.nz/journal LETTER
percentage having ongoing concerns. a virtual clinic as the GP liaison works The importance of the triaging step in closely with the consultant cardiologists of our DAETT pathway is also well demon- the Department. The DAETT service is not strated. The triaging is performed by our GP just providing access to exercise tests and liaison, a general practitioner who works reassurance, it is appropriately targeting a in the Cardiology Department. This results full range of further cardiac appointments, in the DAETT service in effect becoming investigations and management.
Author information: Jessica H Greaves, Medical Registrar; Joan D Leighton, Cardiology GP Liaison; John G Lainchbury, Cardiologist; Paul G Bridgman, Cardiologist. Christchurch Hospital, Christchurch, New Zealand. Corresponding author: Dr Paul Bridgman, Cardiology Department, Christchurch Hospital, Private Bag, Christ- church. (+64 3) 364 0640. [email protected] URL: http://www.nzma.org.nz/journal/read-the-journal/all-issues/2010-2019/ 2015/vol-128-no-1416/6568
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 80 www.nzma.org.nz/journal LETTER
Uptake of new medicines: the Pharmaceutical Management Agency of New Zealand (PHARMAC) in the international context Rajan Ragupathy, Zaheer-Ud-Din Babar
e read Barber and Sheehy’s Institute of Care Excellence (NICE) makes recent analysis of the time the decisions, but does not reprioritise Wfor PHARMAC to fund funding. Reprioritisation is left to regional pharmaceuticals recommended by the fundholding bodies, which are legally Pharmacology and Therapeutics Advisory obliged to implement NICE decisions.3,4 Committee (PTAC) with great interest.1 This approach potentially allows for faster Quantifying the gap between what the uptake of new pharmaceuticals.4 However, health system could ideally provide, and it comes with its own risks. If the health what is actually provided at any given gain from the new pharmaceutical is less time, is vital in debating the trade-offs in than those of a health service that must be the public funding of pharmaceuticals. We foregone to pay for it, then perhaps, the would like to add further to the authors’ result is a net loss to public health.3 discussion of this topic. PHARMAC also differs from its coun- The authors note that the time to fund terparts, such as NICE and the Australian particular pharmaceuticals recommended Pharmaceutical Benefits Advisory by PTAC varies greatly, and is not correlated Committee (PBAC), in that it directly nego- to the clinical priority assigned by PTAC. tiates prices of all pharmaceuticals with However, they do not explore possible manufacturers, and uses this as part of the reasons for these findings, stating “It is funding decision cycle.2 PHARMAC uses a difficult to understand, however, why in variety of techniques to obtain favourable some instances medicines with a lower prices, including competitive tendering, PTAC priority have been funded ahead of generic switches, bundling arrangements those with a high priority”.1 We believe the and therapeutic reference pricing.2 The answer lies in considering the operation opportunities to obtain a favourable of PHARMAC in the international context, price for any given pharmaceutical could and particularly the differences between therefore change due to factors such a ther- PHARMAC and comparable agencies apeutic analogue going off patent, a tender PHARMAC differs from many of its coun- competitor being willing to offer a lower terparts in that it both decides whether price to gain market share, or a bundling a pharmaceutical should be funded and arrangement due to the market entry reprioritises the funding needed.2 No new of an unrelated product from the same pharmaceutical—regardless of its clinical manufacturer. These factors could explain priority—can be funded unless room can be why pharmaceuticals given a lower PTAC found or made within the capped pharma- priority may in fact be funded earlier. ceutical budget to pay for that particular It is also worth remembering that phar- decision. However, this is different in maceutical patents have a limited life, the United Kingdom. There, the National and PHARMAC has in the past been able
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 81 www.nzma.org.nz/journal LETTER
to achieve major price reductions when full discussion of the benefits and drawbacks a given pharmaceutical goes off patent.2 of capped budgets, particularly as these Any externally imposed time constraints have been recently discussed elsewhere.6,7 for PHARMAC to fund a given pharmaceu- We note, however, that Australia is also tical could hamper PHARMAC’s ability to considering a pharmaceutical budget cap.6 find the ‘sweet spot’ in its patent life, and Furthermore, the Australian government has thereby reduce the very efficiency cited by in the past blocked pharmaceutical funding the authors.1 decisions to restrain costs.8 For these reasons, caution is needed when We conclude by noting that every country drawing analogies between PTAC recom- makes often unpalatable trade-offs in mendations awaiting funding, and waiting publically funding pharmaceuticals. New lists for elective services such as surgery. Zealand undoubtedly has delayed access to The latter typically rank patients by clinical pharmaceuticals compared to many other priority, and often set maximum waiting countries.4,9,10 However, the positives in time for each patient.5 This is feasible New Zealand are that it has universal phar- because moving Patient A to the head of maceutical coverage (compared with, for the surgical queue would, all things being example, the United States), and nationally equal, displace ‘only’ Patient B. Moving consistent access for outpatient and hospital Pharmaceutical A prematurely to the head pharmaceuticals (compared with ‘post-code of the funding queue could potentially prescribing’ in the United Kingdom).4 displace funding that could pay for Pharma- Patient cost-sharing for funded pharmaceu- ceuticals B, C and D, thereby resulting in a ticals is lower than many other countries, net loss to public health, despite the indi- and New Zealand also has the certainty of 3 vidual merits of Pharmaceutical A. pharmaceutical cost containment.4,9 Debate The authors correctly note that New about the right balance between these Zealand, unlike Australia, has a capped competing priorities is of course healthy, pharmaceutical budget. Space precludes a and it is a topic we heartily encourage
Competing interests: Nil Author information: Rajan Ragupathy, Clinical Trials and Research Pharmacist, Pharmacy Services, Waikato District Health Board, Hamilton; Zaheer-Ud-Din Babar, Senior Lecturer, School of Pharmacy, Faculty of Medical and Health Sciences, University of Auckland, Auckland. Corresponding author: Rajan Ragupathy, Pharmacy Services, Waikato District Health Board, Pembroke Street, Hamilton 3240, Private Bag 3200. [email protected] URL: http://www.nzma.org.nz/journal/read-the-journal/all-issues/2010-2019/ 2015/vol-128-no-1416/6569
REFERENCES: further research and the 21st Century. Switzer- under single-payer discussion on. land, Adis/Springer 2015. systems in the US, the UK, Australia and New 1. Sheehy K, Barber J. Uptake 3. Barrett A, Roques T, Zealand. Pharmacoeco- of new medicines in New Small M, Smith RD. nomics. 2012; 30: 1051–65. Zealand: evidence of a Rationing: How much will doi:10.2165/11595270- waiting list. N Z Med J. Herceptin really cost? 000000000-00000 2015; 128:1412 10-20 BMJ. 2006; 333: 1118-20. 5. New Zealand Ministry 2. Ragupathy R, Kilpatrick K, 4. Ragupathy R, Aaltonen of Health. Questions and Babar ZUD. Pharmaceutical K, Tordoff J, et al. A answers- elective services. pricing in New Zealand. 3-dimensional view of [Online]. Last Accessed In: Babar Z-U-D, editor. access to licensed and 11 May 2015]. Available Pharmaceutical Prices in subsidized medicines
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 82 www.nzma.org.nz/journal LETTER
at: http://www.health. doi:10.1071/AH14122 of pharmaceutical cost govt.nz/our-work/hospi- 7. Ragupathy R, Babar ZU. containment policies on tals-and-specialist-care/ “Comparing the reimburse- the range of medicines elective-services/ques- ment of new medicines available and subsidized in tions-and-answers-elec- between Australia and Finland and New Zealand. tive-services New Zealand” Aust Health Value Health. 2010; 13: 6. Taylor C, Wonder M. Rev. 2015; (In Press) 148–56. doi:10.1111/j.1524- 4733.2009.00598.x Exploring the implications 8. Shaw B. (2012). Defer- of a fixed budget for new ring PBAC decisions: 10. Wonder M, Milne R. medicines: a study of industry view. Australian Access to new medicines reimbursement of new Prescriber. 2012; 35:3-4 in New Zealand compared medicines in Australia to Australia. N Z Med J. 9. Aaltonen K, Ragupathy R, and New Zealand. 2011; 124:1346 12–28. Tordoff J, et al. The impact Aust Health Rev. 2015;
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 83 www.nzma.org.nz/journal LETTER
Improving our strategy to prevent and control measles outbreaks Lance Gravatt
ike many countries in the Western where they are contacts of measles cases.4 world, New Zealand has been facing The usual MMR vaccination in these cases Lregular outbreaks of measles infection would remain unchanged. with increasing incidence. The Institute of The Pan American Health Organization Environmental Science and Research (ESR) (PAHO) has eliminated measles in South reports that for the 12 months ending Sep- America by applying a measles ‘catch-up’ tember 2014 there were 285 notifications campaign for children >9 months of age.5 of measles over the last 12 months (2013) The Edmonston-Zagreb (EZ) strain of giving a rate of 6.4 cases per 100,000 pop- measles vaccine is the only measles vaccine ulation, a statistically significant increase.1 strain proven to be effective for infants The ESR Report from 2012 states that 55% from the age of 4.5 months6 and to be of measles cases were in children under 3 effective during an outbreak.7 The authors years of age, which matched the fact that of the outbreak immunisation study recom- 55% of confirmed cases had received 0–1 mended the following policy implications of dose of MMR vaccine.2 their research: The Immunisation Sub-Committee of the In situations where control of Pharmacology and Therapeutics Advisory outbreaks is needed, the Edmonston- Committee (PTAC), at its meeting of 23 April Zagreb vaccine could be used to 2013, recommended that the issue of the immunise young household contacts timing of the measles vaccine be revisited if who may have some maternal new data became available to demonstrate antibodies but not enough to prevent long-term efficacy. We present below new them from becoming infected when evidence within the context of existing data exposed at home and to transmit the and consider its implications for measles virus.7 outbreak policy changes in New Zealand. There is some evidence that in infants Those at highest risk of measles aged 4–6 months of age the EZ strain infection are infants under 12 months of is significantly more effective than the age. However, the current Immunisation Schwartz strain in terms of seroconversion Schedule gives the first dose of MMR (62% versus 35%),8 although this study used vaccine at 15 months of age and a booster both subcutaneous and the less effective dose at 4 years of age. intranasal formulations. There was no A 2015 analysis to determine the optimal difference in survival demonstrated at age of measles immunisation, published in 5-year follow-up.9 The NEJM reported the the BMJ, has concluded: results of vaccination of 1,061 6-month old An early two dose schedule at 4–5 infants randomly assigned to either EZ or months and 9 months of age would Schwartz strains of measles vaccine.10 The have been better in terms of reducing seroconversion rate in 6-month-old infants child mortality.3 for standard dose EZ vaccine 18 weeks after This is in line with the current policy of vaccination was 92% compared with 66% the French High Council for Public Health for Schwartz vaccine. from 2013, which recommends infants A 2013 study published in JAMA Paedi- 6–11 months are vaccinated with measles atrics found that if the first dose of MMR
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 84 www.nzma.org.nz/journal LETTER
vaccine was given at 12–24 months of age Authority recognised by the EMA. New there was a significant 2-fold increase in the Zealand has a Mutual Recognition risk of febrile seizure compared to infants Agreement with the European Community who received the measles vaccine only.11 which recognises conformity testing by The study also reported that infants 12–13 Competent Authorities.12 months of age receiving the measles vaccine We submit that measles outbreaks in had roughly half the rate of febrile seizures New Zealand can be better prevented and as the older children in the 19–23 month controlled by including in the National age group. Thus, the risk of febrile seizures Immunisation Schedule: is significantly reduced with a single • Single EZ measles vaccination to measles vaccine given at a younger age infants 4.5 months and 9 months group of infants by as much as 4-fold. For of age with MMR vaccinations this reason alone, a single measles vaccine remaining unchanged at 15 months for infants 4.5 months and 9 months is a preferred option during an outbreak and 4 years compared with MMR vaccine booster. • Single EZ measles vaccination to The EZ measles vaccine is produced from contacts of measles cases where human diploid cells and can therefore be outbreaks occur, including infants as given safely to infants and children who young as 4.5 months of age. are potentially allergic to chicken proteins. New Zealand has the opportunity to effec- The EZ measles vaccine produced by the tively eliminate measles, as South America, Institute of Immunology in Zagreb (IMZ) has by implementing the above strategy. has GMP Certification from the Croatian The global elimination of measles remains Regulator (HALMED) which is a Competent an overdue stated WHO objective.
Competing interests: Nil Author information: Lance Gravatt, Te Arai BioFarma Ltd. [email protected] Corresponding author: Lance Gravatt, Te Arai BioFarma Ltd. [email protected] URL: http://www.nzma.org.nz/journal/read-the-journal/all-issues/2010-2019/ 2015/vol-128-no-1416/6570
REFERENCES: 1. New Zealand Health 5. De Quadros E A et al., sis of a randomised clinical Surveillance Report. Measles Elimination in trial. BMJ. 2008: 337; 1-10. 2014: 12(4); 1-8. the Americas. JAMA. 8. Khanum S et al., Compari- 2. ESR Measles Report. 1996: 275; 224-229. son of Edmonston-Zagreb 2012: March; 1-4. 6. Martins C L et al., A and Schwartz Strains of 3. Aaby P et al., The optimal Randomised Trial of Measles Vaccine Given by age of measles immu- a Standard Dose of Aerosol or Subcutaneous nisation in low-income Edmonston-Zagreb Injection. Lancet. 1997: countries: a secondary Measles Vaccine Given January 17; 150-153. analysis of the assump- at 4.5 Months of Age: 9. Aaby P et al., Long-Term tions underlying the Effect on Total Hospital Survival in Trial of current policy. BMJ Admissions. J Infect Dis. Medium-Titre Edmon- Open. 2012: 2; 1-15. 2014: 209; 1731-8. ston-Zagreb Measles 4. Vaccination against 7. Martins C L et al., Protec- Vaccine in Guinea-Bissau: measles before the age of tive efficacy of standard Five-Year Follow-Up. 12 months. Recommen- Edmonston-Zagreb measles Epidemiol Infect. 1994: dations. High Council of vaccination in infants aged 112; 413-420. Public Health. 2013: July. 4.5 months: interim analy- 10. Markowitz L E et
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 85 www.nzma.org.nz/journal LETTER
al., Immunisation of 11. Rowhani-Rahbar A et al., JAMA Paediatr. 2013: 6-Month Old Infants With Effect of Age on the Risk 167(12); 1111-1117. Different Doses of Edmon- of Fever and Seizures 12. European Medicines ston- Zagreb and Schwartz Following Immunization Agency. National Measles Vaccines. N Engl With Measles-Containing Competent Authorities J Med. 1990: 322; 580-7. Vaccines in Children. (Human). 2015.
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 86 www.nzma.org.nz/journal METHUSELAH
Randomised trial of peanut consumption in infants at risk for peanut allergy The prevalence of peanut allergy among children in Western countries has doubled in the past 10 years, reaching rates of 1.4 to 3.0%, and peanut allergy is becoming apparent in Africa and Asia. As this allergy is the leading cause of anaphylaxis and death due to food allergy, it would be very useful to find means to mitigate the allergy. This trial was conceived to determine whether the early introduction of dietary peanut could serve as an effective primary and secondary strategy for the prevention of peanut allergy. 640 infants with severe eczema, egg allergy, or both, were randomly assigned to consume or avoid peanuts until 60 months of age. The researchers report that the prevalence of peanut allergy at 60 months of age was 13.7% in the avoidance group and 1.9% in the consumption group (p < 0.001). Convincing. N Engl J Med 2015;372:803-13
Latanoprost for open-angle glaucoma Treatments for open-angle glaucoma aim to prevent vision loss through lowering of intra- ocular pressure. The topical application of a prostaglandin analogue is commonly the first choice drug. However, until now there have been no randomised placebo controlled trials to demonstrate the effectiveness of such treatments. This report concerns research aimed to assess the effect of the prostaglandin analogue latanoprost on the visual field preservation of patients with open-angle glaucoma in a comparatively short trial. Ten centres in the UK enrolled over 500 patients who were randomised to receive either latanoprost 0.005% (intervention group) or placebo (control group) eye drops. Drops were administered from identical bottles, once a day, to both eyes. The primary outcome was time to visual field deterioration within 24 months. At the 24 month assessment, the mean reduction in intraocular pressure was greater in the treatment arm of the study. Visual field preservation was found to be significantly longer in the latanoprost patients. Individuals with advanced open-angle glaucoma were excluded so the results of this trial only apply to patients with mild to moderate glaucoma. Lancet 2015; 385:1295-304
Pharmacological treatments for the prevention of acute exacerbations of chronic pulmonary disease This review looks at the clinical trial evidence to support the efficacy of a variety of medications. Trial evidence does support the use of influenza vaccination and pneumococcal vacci- nation. It also supports treatment with long-acting antimuscarinic bronchodilators (LAMAs) and long-acting beta agonist bronchodilators (LABAs) and corticosteroid-LABA combina- tions. However, efficacy for inhaled corticosteroid with LABA and LAMA or LABA and LAMA combination has not been confirmed. Azithromycin has trial evidence backing but the use of pulsed moxifloxacin is not supported. Unsurprisingly, there is no trial evidence backing the use of simvastatin, N-acetyl cysteine or vitamin D. BMJ 2014; 349:g5237
URL: http://www.nzma.org.nz/journal/read-the-journal/all-issues/2010-2019/ 2015/vol-128-no-1416/6571
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 87 www.nzma.org.nz/journal 100 YEARS AGO
The War, the Birth Rate, and Strong Drink
UNDER THE’INFLUENCE. Wife: Good gracious! what do you mean by coming home drunk in broad daylight? Smith: ‘Sorright, dear. People shay I’m ivfluench by prohibition party, an I just been showing ‘em I shupport prohibition of my own accord. (Observer, 13 June 1914). Alexander Turnbull Library, Wellington, New Zealand. http://natlib.govt.nz/records/27579402
he deciding factor which determines the fate of every nation in peace or war is— what will be the numbers of the next generation and who will be their parents?” “TA writer in the New York “Independent” states that in twenty years from now the most serious effects of the war will become apparent. Why is France groaning now under the weight of invasion?—not because she was defeated in 1870, but because she allowed her coffins to outnumber her cradles. The French people of today and of yesterday have sacrificed posterity for luxury and ease and the love of pleasure. They are learning now that it is sacri- fice and not indulgence that is the strength of a nation. Since the year 1881 France has lost a battle a month, as the result of race suicide, and the invasion of her soil began seriously from that period, for foreigners flowed in to fill the vacant places. France was sure of defeat if peace had continued, but providentially she has been given another chance of survival if she will only take heed. It is more honourable for a woman to bear children than for a man to bear arms, and when the function of women is exalted in a nation, and morality is exalted there is less risk of war, and if war should arise, every hope of its being carried to a successful issue.
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 88 www.nzma.org.nz/journal 100 YEARS AGO
France has been the great sinner in limiting the birthrate, but even our own little country has transgressed in a less degree. The summum bonum of many women here has been ease and amusement, and the offenders are those women who belong to the most prosperous classes of society. The wives of professional men do not, we believe, have a family, on the average, of more than two or three children. The birth-rate in New Zealand per 1,000 of popu- lation in 1882–86 was 35.4; in 1899 it had dropped to 25, and in 1913 had risen only to 26. The price of this state of affairs to be paid by New Zealand is that we are not able to send as many recruits to the war as we ought to have been able to send, and the wealth of this country is less than it ought to be. In 1878 one in every three of the married women of childbearing age in New Zealand gave birth to a child, while in 1911 the rate was only one in nearly five. Germany has not perpetrated these national crimes. The Germans had not until recently adopted a policy of race-suicide, nor did they make pleasure the great goal of life, but they offended even far more seriously in the sins of the spirit, pride, greed, envy, self-sufficiency, and unbelief in matters that are supreme and eternal. The decline in the birth-rate of the British Isles is also deplorable, but the greatest curse of England is the appalling lust for strong drink. It is almost impossible to exaggerate this factor in the production of insanity, crime, poverty, misery; disease, and national inefficiency. We have all read reports from statesmen, soldiers of high standing, and the Admiralty of the effect of drink in lessening the supply of munitions and warships for the prosecution of this great war that is being fought for freedom. Everyone who has lived in the great cities at Home has observed the ravages of the drink fiend, and here in New Zealand we see the same destruction in less degree. Oh, God of justice! God of right! Why is the world so full of woe? Why are souls withered by this blight? ls this the working of Thy foe— The rebel sire of sin and crime, Who makes Thy likeness, man, his thrall? Oh, Father! Shield our new-born clime From this, the vilest fiend of all. Bracken, who wrote these lines, was neither a Puritan nor a Prohibitionist. We do not believe that the natural appetite for stimulants is wrong, or that facilities for gratifying it can be entirely abolished, but we do believe that sharp measures for restricting the lust for alcohol in the British race should be attempted, if only for the sake of efficiency in war. If these measures fail, we shall welcome an attempt at prohibition, for it is better that the moderate man should go without than that in a time of stress the immoderate should materially lessen the stability of the army, the navy, and the nation. Mr. Lloyd George, although, we disagree with him on many points, in this matter of the drink curse has earned the thanks of the nation. He has attempted to lessen the facilities for drunkenness, and to provide beverages less loaded with alcohol than those which gratify the drunkard’s palate. He has had vested interests against him. The Irish Nationalists, who help to keep him in office, backed by Irish publicans and distillers, have warned him to take his sacri- legious hands off the Ark of the Covenant, which is, in this instance, the whisky cask. It is better for the nation to suffer than that the profits of the trade should dwindle, or that the brewers should alter the quality of their output! The House of Lords contains dozens of peers directly and financially interested in the liquor trade. The Bench of Bishops, so far as we know, has not seriously attempted to lessen the facilities for drunkenness. It is time for the medical profession to give its support to the cause of temperance; and we recognise that we cannot stop drinking but we can lessen drunkenness by condemning the most ardent liquors, and checking facilities for the sale of drink, and generally lessening temptation. In other words, let us not lead moral weaklings into temptation but deliver them from evil. Editorial, NZMJ June 1915
URL: http://www.nzma.org.nz/journal/read-the-journal/all-issues/2010-2019/ 2015/vol-128-no-1416/6572
NZMJ 12 June 2015, Vol 128 No 1416 ISSN 1175-8716 © NZMA 89 www.nzma.org.nz/journal