HIV Prevention Today: with Coordinated Action, We Can End Transmission Infrequent Condom Use with Casual Partners Among New Zealand Gay and Bisexual Men

Journal of the New Zealand Medical Association Vol 128 | No 1426 | 4 December 2015 HIV prevention today: with coordinated action, we can end transmission Infrequent condom use with casual partners among New Zealand gay and bisexual men

Drug misuse in sport

The first 30 years of HIV Maxillofacial fractures Late-life self-harm in in New Zealand: Review at Waikato Hospital, the Waikato region of the epidemiology New Zealand: 2004 to 2013 What makes a child a ‘competent’ child? 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 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 2 www.nzma.org.nz/journal CONTENTS

EDITORIAL 75 8 Late-life self-harm in the HIV prevention today: with Waikato region coordinated action, we can end WA de Beer, J Murtagh, G Cheung transmission 83 Peter J Saxton, Anthony J Hughes, Prescriber compliance with Massimo Giola renal function monitoring 16 in patients taking dabigatran Drug misuse in sport: a (Pradaxa) historical perspective Katie Thorne, Stephen Dee, Sisira David Gerrard Jayathissa ARTICLES 88 What makes a child a 19 ‘competent’ child? Thirty years of condom-based Amanda van Rooyen, Tineke Water, HIV prevention by gay men in Shayne Rasmussen, Kate Diesfeld New Zealand Anthony J Hughes, Peter J Saxton 96 Maxillofacial fractures at 31 Waikato Hospital, New Zealand: The first 30 years of HIV in 2004 to 2013 New Zealand: Review of the Blake K Moore, Ryan B Smit, Angus N epidemiology Colquhoun, W Murray Thomson Nigel Dickson, Bible Lee, Timothy Foster, Peter Saxton VIEWPOINT 49 103 Infrequent condom use with A bug in the ointment: topical casual partners among New antimicrobial usage and Zealand gay and bisexual men resistance in New Zealand Peter J Saxton, Nigel P Dickson, Deborah A Williamson, Stephen R Anthony J Hughes, Adrian H Ludlam Ritchie, Emma Best, Arlo Upton, Alison Leversha, Alesha Smith, 62 Mark G Thomas Drug misuse in sport: a CLINICAL CORRESPONDENCE New Zealand perspective Andrew Curtis, David Gerrard, Peter 110 Burt, Hamish Osborne Successful conservative 69 management of campylobacter ‘Real-time’ burden of community cholecystitis occurring post and healthcare-related infections chemotherapy and rituximab: a in medical and rehabilitation rare disease entity patients in a public hospital in Ajay Gupta, Louise Teo Auckland, New Zealand Kerry Read, Hasan Bhally

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 3 www.nzma.org.nz/journal CONTENTS

LETTERS METHUSELAH 113 119 Over five years on, we still don’t need a pilot bowel cancer 100 YEARS AGO screening programme: Please 120 just get on with it! Post mortem examinations a Guy Hingston public service 116 PROCEEDINGS Modelling of tobacco endgame interventions: a response 122 Richard Edwards, Tony Blakely, Selected proceedings of the Frederieke van der Deen APAC Forum 2015

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 4 www.nzma.org.nz/journal SUMMARIES

Infrequent condom use with casual partners among New Zealand gay and bisexual men Peter J Saxton, Nigel P Dickson, Anthony J Hughes, Adrian H Ludlam In 2014, the highest number of gay men were diagnosed with HIV in New Zealand. Our study examined condom use in a large and diverse sample of this population. Of those having intercourse with casual partners, a quarter (27.2%) reported infrequent condom use. Attitudes to condoms and exposure to condom promotion predicted condom use, among other factors such as adoption of condoms early on in life. Internet dating sites are common places to meet sexual partners. The relationship these commercial sites have with HIV prevention and public health agencies is mixed. Cooperative models where dating sites are partners in prevention should be developed. HIV and many STIs are asymptomatic communicable infections. Like vaccination, high levels of condom use are needed to control transmission within communities. The increases in HIV and STI diagnoses seen in New Zealand provide early warning of riskier behaviours happening.

Thirty years of condom-based HIV prevention by gay men in New Zealand Anthony J Hughes, Peter J Saxton Three decades after the first government-funded HIV prevention campaign in 1985, gay and bisexual men remain the population most at risk of infection in New Zealand. We review the major reasons why this is the case and describe the approach that has been taken to HIV prevention here. We have a good international record in HIV prevention for gay men, however new HIV diagnoses rates are increasing once again around the world. Lessons from the first three decades must underpin future HIV control efforts and additional intervention strategies will also now be required.

The first 30 years of HIV in New Zealand: Review of the epidemiology Nigel Dickson, Bible Lee, Timothy Foster, Peter Saxton New Zealand has a well-described low prevalence epidemic of HIV infection, mostly concentrated in sub-populations of men who have sex with men (MSM), and heterosexual individuals from sub-Saharan Africa and South-East Asia. The former is largely due to transmission within New Zealand, whereas the latter mostly occurred overseas, although the difference has been less marked in recent years. The number of notified cases of AIDS peaked in the late 1980s, and dropped dramatically in the mid-1990s due to the introduction of effective antiretroviral treatments. Presently most cases of AIDS are in people with previously undiagnosed HIV infection. In contrast, currently the annual number of diagnoses of HIV infection is higher than in the late 1990s due to more occurring among MSM. Over the past 30 years each sub-epidemic has demonstrated a distinct pattern reflecting different determinants. Control of HIV in New Zealand is favourable compared to many countries, however challenges remain, especially in prevention among MSM and more timely diagnosis for all, especially those heterosexually infected.

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 5 www.nzma.org.nz/journal SUMMARIES

Drug misuse in sport: a New Zealand perspective Andrew Curtis, David Gerrard, Peter Burt, Hamish Osborne Drug misuse in sport is an international phenomenon that has not escaped the attention of health professionals in New Zealand. Young athletes are vulnerable to the use of performance-enhancing substances and the increasing use of sports supplements reflects a particularly worrying trend fostered by unscientific endorsements. Drug-Free Sport New Zealand is the national anti-doping agency responsible for the oversight and education of our athletes in an environment where sport is an integral part of our culture. Doctors responsible for the care of athletes have an obligation to respect the Code of the World Anti-Doping Agency.

What makes a child a ‘competent’ child? Amanda van Rooyen, Tineke Water, Shayne Rasmussen, Kate Diesfeld To give informed consent to healthcare in New Zealand, competence is a requirement. A person needs to understand the nature, purpose and consequences of treatment and non-treatment in order to give a legally valid agreement to healthcare. However, New Zealand law is unclear on this matter where children are concerned. Although not overtly stated, New Zealand law infers that children under the age of 16 years may give or withhold consent to healthcare, independent of their parents, so long as they are competent to do so. This article raises the questions; what is ‘child competence’, why is it so important to acknowledge and how do healthcare professionals assess for child competence? Unfortunately, there is meagre research in this area and no clear answers. The assessment, recognition and respect for a child’s level of competence not only supports ethical arguments regarding respect for their rights and personhood; it has other more tangible benefits to both the child and healthcare services. These include improved treatment adherence, clinical effectiveness, health service delivery and disease prevention. Therefore, this article addresses how these benefits can be realised through a better understanding and assessment of children’s abilities to participate in and consent to healthcare.

Maxillofacial fractures at Waikato Hospital, New Zealand: 2004 to 2013 Blake K Moore, Ryan B Smit, Angus N Colquhoun, W Murray Thomson The rate of facial fractures presenting to Waikato Hospital has been increasing since 1989. In addition the rate of violence-related facial fractures is now at almost double the rate seen in 1998-2000. It continues to be the dominant cause of injury, while road traffic accident related fractures are decreasing. This continual increase in fractures presenting to Waikato Hospital places significant demands on scarce clinical resources, such as operating theatre time and staffing numbers. Violence is an escalating cause of facial fractures that requires urgent and interventional public health prevention strategies.

A bug in the ointment: topical antimicrobial usage and resistance in New Zealand Deborah A Williamson, Stephen R Ritchie, Emma Best, Arlo Upton, Alison Leversha, Alesha Smith, Mark G Thomas New Zealand has extremely high rates of bacterial resistance to topical antibiotics such as Bactroban and Foban. This is because we use a lot of these antibiotics. In this article, we look at who gets prescribed topical antibiotics in NZ, and suggest some ways in wihch we might control and reduce rates of resistance.

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 6 www.nzma.org.nz/journal SUMMARIES

‘Real-time’ burden of community and healthcare-related infections in medical and rehabilitation patients in a public hospital in Auckland, New Zealand Kerry Read, Hasan Bhally This study, performed in a single public hospital in Auckland which does not provide acute surgical services, shows that a high proportion of patients admitted to hospital have a diagnosis of infection as the main or a secondary reason requiring admission- the common types being chest, urine and skin infections. It also confirms the previous finding that hospital acquired infections are relatively common with approximately 1 in 10 patients acquiring it during their hospital stay.

Late-life self-harm in the Waikato Region WA de Beer, J Murtagh, G Cheung This paper looked at the attributes of late life suicide attempts i.e. in the group over 65 years of age, in the Waikato region. Elderly people identified that concomitant medical conditions and depressive illnesses were important stresses contributing to their self-harm attempts. The study highlighted some important strategies for identifying and better managing elderly people who may attempt suicide in the future.

Prescriber compliance with renal function monitoring in patients taking dabigatran (Pradaxa) Katie Thorne, Stephen Dee, Sisira Jayathissa Dabigatran, a novel ani-coagulant medication, has been licensed for use in New Zealand since 2011. It is a drug that is mainly cleared by the kidneys and if the kidney function becomes impaired there is an increased risk of bleeding. Current guidelines advise checking kidney function before prescribing dabigatran and at least once a year following this. This study looked at whether patients had their kidney function checked in accordance with these guidelines. We identified one third of patients did not have their kidney function checked at least once a year. Many of these patients were elderly and or had pre-existing kidney impairment. We advise that further measures should be implemented to improve kidney function testing. One method might be to introduce an automated electronic reminder.

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 7 www.nzma.org.nz/journal EDITORIAL

HIV prevention today: with coordinated action, we can end transmission Peter J Saxton, Anthony J Hughes, Massimo Giola

ew Zealand has an enviable inter- and be the first country to do so. This will national record in HIV prevention, take revived political will and adequate Nwith diagnosis rates for most-at- resourcing. While 2014 brought harbingers risk groups being among the lowest in the of a worsening epidemic phase, scientific world.1 HIV infection is now treatable, breakthroughs this year, 2015, gave us the affected communities have established tools to constrict transmission through multi- cultures of risk reduction,2,3 and laws and pronged interventions if we respond boldly. policies have mostly been aligned to sup- We call this approach ‘comprehensive port prevention.3 prevention’, and five locally relevant action Successes so far cannot be allowed to points are summarised in Table 1. engender complacency. HIV transmission First, condom-based HIV prevention for is preventable, but not declining; an HIV GBM, who remain most at risk in this country, diagnosis is still life changing, and HIV must continue and become even more sophis- medication is expensive. Government ticated. This is because barrier protection funding for antiretroviral therapy (ART) responds so well to qualities of HIV trans- alone has risen from $14.5 million in mission during sexual behaviour, because it mid-2010 to $26.4 million by mid-2014 is cheap, and because it is easily scaleable.3 (prior to PHARMAC discounts) for a rela- Second and third, we must deploy the full tively small number of patients.4 repertoire of treatment-based prevention Warning signs of a reversal in control interventions, particularly immediate ART in New Zealand are becoming apparent. on diagnosis to minimise HIV transmission Last year 117 gay and bisexual men (GBM) risks from those infected, and pre-exposure were newly diagnosed with HIV infection— prophylaxis (PrEP) to minimise HIV acqui- including 86 who were infected here—the sition risks for the subset of uninfected highest annual number ever recorded.1 The individuals at highest risk of exposure. proportion of GBM engaging in unprotected Fourth, HIV testing access and frequency casual sex, although low by international must be improved to provide timely pathways standards, increased in 2014.2 Infectious into these twin treatment-based prevention syphilis cases among GBM reported by sexual levers. Fifth, vaccination against sexually health clinics doubled in 2014 in some regions transmitted infections (STI) such as HPV, including Auckland, and the number of rectal HAV and HBV needs to be expanded to all gonorrhoea cases reported in males rose from at risk groups, and screening and treatment 31 in 2010, to 121 in 2014;5 both are proxies practices must be enhanced, because of the for changes in risky sexual behaviour. The synergies between STI and HIV control. small number of HIV infections acquired Together, these action points synchronise heterosexually in New Zealand is incre- condom-based and treatment-based HIV 1 mentally rising. In many countries, we are prevention strategies to reduce the repro- watching HIV prevention in GBM unrav- ductive rate of HIV below replacement. elling, spurring an urgency to adapt our own The possibilities of this approach were responses now, before the achievements of recently modelled in Australia, where a the past 30 years are squandered. 44% reduction in HIV diagnoses nationally If we do adapt quickly we can virtually was estimated in the first year if condoms, eliminate HIV transmission in New Zealand, ART and PrEP were mobilised in tandem

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 8 www.nzma.org.nz/journal EDITORIAL

Table 1: Five actions to eliminate HIV transmission in New Zealand

Action Goal (1) Increasingly sophisticated promotion of To interrupt HIV and STI transmission condoms for protection against HIV and STIs during anal and vaginal intercourse

(2) Immediate access to HIV antiretroviral To minimise infectiousness and maximise treatment post diagnosis personal wellbeing of individuals with confirmed HIV infection

(3) A Government-funded programme of To target the most at risk individuals who play a voluntary pre-exposure prophylaxis (PrEP) disproportionate role in fuelling the HIV epidemic and quarterly STI screening for the minority of high risk individuals who are unable to sustain consistent and correct condom use during anal and vaginal intercourse

(4) Prompt HIV testing following intercourse To reduce the number with undiagnosed HIV without a condom by rapid testing and infection potentially home HIV testing or home HIV sampling

(5) Improved access to comprehensive STI To control resurgent STI epidemics that synergise screening, treatment and vaccination with HIV control at sufficient scale among GBM.6 There is Adding ‘test and treat’ approaches government enthusiasm to achieve this to condoms and behaviour change can across the Tasman; New Zealand cannot dramatically alter the trajectory of the HIV afford to merely watch and wait. epidemic in New Zealand; we discuss some Fortunately, New Zealand has an excellent opportunities and challenges. platform to launch a new phase in our epidemic response, being a low HIV prev- Opportunities alence country with a committed HIV Treatment for confirmed HIV workforce and enjoying the small scale that enables coordinated action. But, we will infection In July 2015, two studies confirmed that also need more allies. Regulatory change early treatment of HIV infection is critical is urgently needed to remove PHARMAC’s for optimising long term health.10,11 Notably, CD4 prescribing obstacle for ART for indi- the START study reported that the risk of viduals with confirmed HIV infection, and bring it into line with new WHO guide- developing serious illness or death was lines.7 Protocols for prescribing targeted 57% lower among those treated early when PrEP to uninfected high-risk individuals CD4+ counts were above 500 cells/mm3, need to be accelerated.8 Funding for NGOs compared to those in the deferred group at the frontline should match the increased treated when their CD4+ counts declined 3 10 capacity needed to add testing and PrEP below 350 cells/mm . Early ART initi- and STI screening promotion onto condom ation also showed minimal or no increase 12 13 promotion. Research, surveillance and in adverse events. The WHO and other 14 evaluation of the HIV sector’s performance international bodies now recommend must be resourced on a sustainable basis. access to ART for all individuals with Inadequate responsiveness to health issues confirmed HIV infection regardless of CD4+ affecting GBM must be corrected, and count, and New Zealand must urgently greater inclusivity of sexual orientation follow suit. minorities fostered.9 And the ongoing stigma In 2015, the prevention benefits of ART surrounding people living with HIV must were also confirmed. In final results from be stridently challenged—while not dimin- a large trial of HIV discordant, mostly ishing the threat that HIV poses—if we heterosexual couples, ART conferred a are to motivate individuals to engage with 93% protective effect on transmission.15 enhanced HIV screening and care options. Interim analyses of two further cohort

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 9 www.nzma.org.nz/journal EDITORIAL

studies in 2014 and 2015 found no HIV HIV testing transmissions between HIV discordant HIV testing has to improve to fully realise gay male couples where the HIV infected the public health benefits of HIV treat- 16,17 partner had an undetectable viral load. ments at the population level, as neither Although a low rate of transmission ART nor PrEP can be offered in absence of cannot be ruled out, ART undoubtedly confirmed HIV status. Of paramount impor- offers a strong preventive effect at the tance is reducing the number of individuals individual level with full adherence. with undiagnosed infection, estimated to HIV pre-exposure prophylaxis (PrEP) be around 600 in New Zealand,1 based on PrEP is the daily use of HIV antiretroviral a 2011 study that found 1 in 5 HIV positive 25 drugs, such as Truvada, by confirmed HIV GBM were unaware they were infected. negative individuals to reduce the like- Currently, 42% of GBM report having 26 lihood of HIV acquisition.13 PrEP should be tested in the last year. A 2015 preliminary taken 5–7 days prior to anal intercourse to study modelled that the estimated 10.3% achieve optimum protective concentrations of Australia’s HIV infected population who and adherence is key for lasting impact. In remain undiagnosed contributes 44.9% 27 the iPrEx trial among GBM, PrEP was 44% of onward HIV transmissions. Shifting protective overall, with an adherence-ad- undiagnosed individuals into clinical justed protection of 92%.18 Taking tablets management and reducing their viral load 4 or more days a week appears to achieve through ART as soon as possible is conse- full protection.19 Interim results of two quently a high priority for HIV control as subsequent clinic-based studies in 2015 well as for their own care. Testing access, (PROUD20 and Ipergay21) both reported an testing uptake, testing frequency and explo- overall 86% protective effect. ration of new testing technologies, such as home testing or home sampling, therefore WHO has now recommended PrEP in need to be key targets in the new HIV populations where HIV incidence exceeds prevention era. 3 per 100py,13 and Australia has several demonstration projects underway among high risk GBM. This includes individuals Challenges reporting any of the following in the last ART: Matching promise with 3 months and who are likely to continue practice this behaviour in the next 3 months: Against the undisputed prevention i) an HIV positive partner with whom promise of treating HIV positive individuals, condoms were not used; ii) receptive anal ‘test and treat’ approaches have to overcome intercourse with a casual male partner several hurdles. HIV testing is essential to of HIV positive or unknown status; iii) a identify those infected, and increases both bacterial rectal STI; iv) methamphetamine in coverage and frequency can reduce HIV use.6 Participants in most projects inter- incidence.28 However, even more frequent nationally represent a high-risk subset testing of all GBM may not intervene soon of GBM with low rates of condom use, enough to halt the epidemic, because a small multiple partners, and high baseline and proportion of individuals typically drive incident STIs.22 Preliminary data suggest clusters of transmission,29 the gap length that in a high HIV incidence setting, such between sexual partners for these men as the UK, temporary and targeted PrEP to can be very brief, and a large fraction of high-risk individuals can be cost effective, transmissions therefore occur in the highly or even cost saving, when compared infectious period following HIV infection, to the cost of lifelong HIV treatment of but prior to testing.30 Even in societies with people whose infection could have been high rates of HIV testing, such as GBM in averted, although this depended heavily on Australia,31 the median time from infection assumptions about uptake and future drug to diagnosis is estimated to be 1.4 years,32 discounts.23,24 New Zealand needs local data implying that those who test often are not describing anticipated PrEP uptake so that always those most at risk. In New Zealand, the same cost/benefit calculations and eligi- over a third of GBM newly diagnosed are bility can be assessed. diagnosed late,1 and around half of GBM

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 10 www.nzma.org.nz/journal EDITORIAL

engaging in risky sex in the last six months primary prevention. Undoubtedly, the best had not tested since that episode.26 Increased outcomes will be achieved when condom- HIV testing promotion and newer HIV based and treatment-based HIV prevention testing technologies should be responsive to are mutually reinforcing and do not the needs of these individuals first. undermine each other. For all these reasons, Then a number of milestones need to be increased promotion of HIV testing and early met, such as linking and retaining those treatment must not erode condom advocacy. diagnosed into specialist care, ensuring PrEP: uptake, targeting and equity treatment access and adherence to medi- Likewise, we need to motivate the most cation, and sustaining undetectable viral at risk individuals to attend medical clinics load. Attrition occurs at each step in this if we are to fully capitalise on PrEP. PrEP “HIV-care cascade”,33 and no country to is a programme, not simply a prescription, date has achieved the UNAIDS target of 73% involving a high level of clinical monitoring, of HIV infected individuals having unde- including regular HIV and STI screening, tectable viral load,34 although cities such as drug adherence and adverse events, and Stockholm provide a model to emulate.35 ongoing safe-sex counselling. Cost-effec- Maintaining access to a range of affordable, tiveness will vary by setting and eligibility effective and tolerable HIV medication as criteria, being influenced by background trade agreements are negotiated will also be HIV incidence and local healthcare costs.23,24 an important government goal in order to It is unclear whether PrEP is a temporary or meet the needs of all HIV patients. long-term option for some individuals, and Modelling from the UK in 2015 helps under what circumstances it will be ceased. us visualise possible targets.36 Achieving Affordability is a pressing issue, as Truvada 90% of all HIV infected GBM diagnosed, is currently an expensive on-patent medi- treated and undetectable within one year cation. Targeting will be necessary to of infection would require a more than minimise PrEP uptake among low-risk trebling in the annual number of HIV individuals, conserve drug stocks, and avoid tests conducted (to 65% of all GBM tested drug resistance developing. annually), ART being initiated on diag- There are also concerns about reduc- nosis, and retention in care and treatment tions in community-wide condom use, and adherence remaining high and not increases in STIs, resulting from percep- declining. This would reduce HIV inci- tions that the combination of PrEP and dence by 79% by 2030, push the epidemic ‘test and treat’ have already eliminated below replacement long term, and be cost HIV acquisition risks, regardless of the effective. Each additional 10% of HIV-in- actual level of population scale up. Scant fected GBM virally suppressed from the consideration has so far been given to current situation of 58% equated to 37% equity, despite evidence of uneven access fewer HIV infections per year. to HIV services within GBM communities These potential epidemiological gains particularly by ethnicity,33 and low rates from ‘test and treat’ can be counteracted of sexual orientation disclosure,9 both of if condom use deteriorates. If the opti- which are required before PrEP can be mistic rhetoric surrounding HIV treatments offered. Unlike a condom, PreP is a pill reduces condom use by a mere 10%, and that has to be taken for several days prior yet HIV testing levels and ART initiation to sex in order to build up protective levels, do not improve, HIV incidence is predicted making it an unverifiable intervention for to double in 15 years.36 Increases in casual sex partners. This raises important condomless anal sex can trigger resurgent issues of power asymmetry during sexual STI epidemics that are serious in their own encounters, as it relies on people commu- right and heighten HIV transmission and nicating honestly and without coercion in acquisition risks. Moreover, a singular focus the heat of the moment. Nevertheless, if on HIV testing as the entry point into HIV tightly targeted to the most at risk and moti- prevention is also arguably unethical, as it vated individuals, PrEP has considerable will have delayed intervention until after potential to improve HIV control by inter- infection has already happened, neglecting rupting chains of transmission in the most opportunities for earlier condom-based vulnerable subsets of the community.

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 11 www.nzma.org.nz/journal EDITORIAL

Accommodating both public Community-based advocates in response health and clinical medicine have pointed to the gradual erosion of an already imbalanced funding quota and The championing of pharmaceuticals the inevitably diminished population-level for HIV prevention, as well as for HIV care, impact of their work. Economic analysis has shifted the momentum internationally commissioned by NZAF suggests that towards clinic-based HIV control. This investment in HIV risk behaviour change emerged after 1996 with the effectiveness of is cost effective,41 but primary prevention the first triple combination ART and hastened will only succeed if delivered at suffi- following the influential clinical trial results cient intensity. And while the substantial from the HPTN 052,15 iPrex18 and PARTNER16 investment in HIV clinical medicine has trans- studies in the last 5 years. However, some formed life expectancy for people living with public health practitioners have challenged HIV, and treatment has high individual-level the privileging of such trials as the gold prevention efficacy, public health researchers standard for scientific decision making about have noted that pharmaceuticals have not interventions in real-world communities. so far been a panacea for controlling HIV at Shelton, for example, has argued that the the population level.42 Given finite resources, issue is not only whether an intervention analyses remind us that investment in (such as HIV treatment) is efficacious in an community-based education is still more RCT, but if it can be made to work practicably cost-effective than ART, and ART is more at scale, be additive rather than zero-sum cost-effective than PrEP.43 against other interventions, and be sustained over long periods given available resources Medicalisation of HIV prevention and capacity.37,3 Situational variability These shifts have heralded a growing means that successful interventions may not medicalisation of HIV prevention. Some translate faithfully elsewhere. In this view, social researchers have argued that clin- ‘what works’ is deducing the optimal combi- ic-based prevention models privatise nation of interventions for local conditions.37 HIV and remove it as a subject of public Research on New Zealand-based experiences discussion, debate and action—the latter would sharpen implementation and help being foundations of the early, effective, 42 avoid unintended effects. community-based HIV response. Medi- calisation also tends to position GBM as Avoiding disinvestment in patients and consumers of diagnostic tech- community-based HIV prevention nologies and pharmaceuticals. Those may Disinvestment in behaviour-based be easier identities for the health system to HIV prevention programmes has corre- engage with, but it can bypass the topic of spondingly become a concerning trend sex and sexuality, and sterilises discussions internationally. In the mid-1990s, a of HIV prevention. In doing so, medical- contraction in high-level investment and isation de-emphasises the agency of gay coordination in several Australian juris- men and other groups to safeguard their dictions, but not in New South Wales, own sexual health, and avoids the need preceded an increase in HIV notifications.38 to address vulnerabilities driven by social In England, GBP 2.9 million was spent marginalisation. on national HIV prevention programmes Furthermore, difficulties achieving in 2011/12 and reducing, less than half a the necessary scale, clinical linkage and percent of the GBP 762 million spent on retention, medication affordability, access treatment and care and rising.39 Since 2011, and adherence have to date limited the government funding for the New Zealand full public impact of pharmaceutical-based AIDS Foundation (NZAF) has been static prevention interventions.31 These are the (around $4.2 million),40 while HIV treatment same criticisms often levelled against expenditure here has risen 57% to $26.4 behaviour-based programmes: both require million.4 The reallocation of HIV funding repeated actions to be effective (testing portfolios towards clinical services overseas and taking medications regularly; using has typically been justified by claims of condoms consistently). Unsurprisingly, condom-based ‘prevention failure’, because some individuals are reluctant to have their HIV diagnosis rates have not declined. sexual lives revolve around clinic appoint-

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 12 www.nzma.org.nz/journal EDITORIAL

ments and medication, in much the same (‘test and treat’). We must deliver this in a way that some men discontinue regular way that maximises, not compromises, the condom use. effectiveness of each of these approaches. This recognises that effective clinic-based Conclusion prevention is dependent on vigorous HIV infection is avoidable and unnec- community-based promotion of testing; essary. Today’s prevention tools offer the similarly, effective community-based possibility of virtually eliminating HIV condom promotion must be supported by transmission if we can re-activate the spirit advocacy for continued condom use in the of cooperation that defined New Zealand’s clinic setting. Once such a programme is early successes. This task now calls for established, the most challenging task will a partnership for prevention: a strategic then be to maintain it over long periods cooperation between community-based of time, galvanising ongoing political primary prevention (condom promotion), commitment and community engagement, clinic-based primary prevention (PrEP)44, until a sterilising vaccine or cure eventually and clinic-based secondary prevention becomes available.

Competing interests: Peter Saxton is the current New Zealand AIDS Foundation Fellow at the Faculty of Medicine, University of Auckland. Tony Hughes is Scientific Director at the New Zealand AIDS Founda- tion. Massimo Giola serves on the Board of Trustees of the New Zealand AIDS Foundation. Author information: Peter J W Saxton, Director, Gay Men’s Sexual Health research group, Department of Social and Community Health, University of Auckland, Auckland; Anthony J Hughes, Scientific Director, New Zealand AIDS Foundation, Auckland; Massimo Giola, Consultant Physician, Bay of Plenty District Health Board, Tauranga Hospital, Tauranga. Corresponding author: Dr Peter Saxton, Department of Social and Community Health, University of Auckland, Private Bag 92109, Auckland. [email protected] URL: www.nzma.org.nz/journal/read-the-journal/all-issues/2010-2019/2015/vol-128-no-1426-4- december-2015/6742

REFERENCES: 1. Dickson NP, Lee SKB, all-issues/2010-2019/2015/ esr.cri.nz/surveillance/ Foster T, Saxton PJ. The vol-128-no-1426-4-de- surveillance.php first 30 years of HIV in cember-2015/6746 6. Zablotska I. The Australian New Zealand: Review 3. Hughes AJ, Saxton PJ. National PrEP guide- of the epidemiology. NZ Thirty years of condom- lines and the potential Med J. 2015 128:1426. based HIV prevention by impact of PrEP on HIV http://www.nzma.org.nz/ gay men in New Zealand. prevention in Australia. journal/read-the-journal/ NZ Med J. 2015 online first. Presented at Australasian all-issues/2010-2019/2015/ 4. Ministry of Health. Coun- HIV&AIDS Conference, vol-128-no-1426-4-de- try Progress Report New 16-18 Sept 2015, Brisbane. cember-2015/6745 Zealand (HIV/AIDS) 2014. 7. New Zealand Sexual 2. Saxton PJ, Dickson NP, Wellington: Ministry of Health Society. Hughes AJ, Ludlam AH. Health, 2015. http://www. Expanded access to HIV Infrequent condom use health.govt.nz/system/files/ antiretroviral therapy with casual partners documents/publications/ for all individuals with among New Zealand gay nz_country_progress_ confirmed HIV infection. and bisexual men. NZ report_jan_dec_2014.pdf. Submission to Pharmac, Med J. 2015 128:1426. 5. The Institute of Envi- 20 Nov 2015. https://www. http://www.nzma.org.nz/ ronmental Science and fmhs.auckland.ac.nz/ journal/read-the-journal/ Research. https://surv. assets/fmhs/soph/sch/

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 13 www.nzma.org.nz/journal EDITORIAL

gmsh/docs/NZSHS%20 lines.aspx; European AIDS a pragmatic open-label Application%20to%20 Clinical Society http:// randomised trial. Lancet. Pharmac%20HIV%20 www.eacsociety.org/ 2015; online first http:// antiretroviral%20 guidelines/eacs-guidelines/ dx.doi.org/10.1016/ initiation%2020%20 eacs-guidelines.html S0140-6736(15)00056-2 11%2015.pdf. 15. Cohen M, Chen Y, McCau- 21. Molina JM, Capitant C, 8. GayNZ.com. “HIV ley M, et al. Final results Spire B, et al. On Demand prevention: we risk being of the HPTN 052 random- PrEP With Oral TDF-FTC left behind”. 22 Sept ized controlled trial: in MSM: Results of the 2015. http://www.gaynz. antiretroviral therapy ANRS Ipergay Trial. com/articles/publish/7/ prevents HIV transmis- Presented at Conference article_17328.php sion. J Int AIDS Soc. on Retroviruses and 9. Ludlam AH, Saxton PJ, 2015;18(5Suppl. 4): 20479. Opportunistic Infections, Dickson NP, Hughes AJ. 16. Rodger A, Bruun T, 23-26 Feb 2015, Seattle. General practitioner Cambiano V, et al. HIV 22. Volk JE, Marcus JL, awareness of sexual orien- transmission risk through Phengrasamy T, et al. No tation among a community condomless sex if HIV+ new HIV infections with and internet sample of gay partner on suppressive increasing use of HIV and bisexual men in New ART: PARTNER Study. preexposure prophylaxis Zealand. J Prim Health Presented at Conference in a clinical practice Care. 2015;7:204-12. on Retroviruses and setting. Clin Infect Dis. 10. The INSIGHT START Opportunistic Infec- 2015; 61:1601-3. Study Group. Initiation tions, 3-6 March 2014, 23. Cambiano V, Miners A, of antiretroviral therapy Boston. Oral late breaker Dunn D, et al. Is PrEP for in early asymptomatic abstract 153LB. HIV prevention cost-ef- HIV infection. N Engl J 17. Grulich AE, Bavinton BR, fective in MSM in the Med. 2015; 373:795-807. Jin F et al on behalf of the UK? Presented at British 11. The TEMPRANO ANRS Opposites Attract Study Association for HIV and 12136 Study Group. A trial Group. HIV Transmission Sexual Health Conference, of early antiretrovirals in Male Serodiscordant 1-3 June 2015, Glasgow. and isoniazid preventive Couples in Australia, Thai- 24. Ong K, Desai S, Desai M et therapy in Africa. N Engl land and Brazil. Presented al. The cost-effectiveness J Med. 2015; 373:808-22. at Conference on of Pre-Exposure Prophy- Retroviruses and Oppor- 12. Abdool Karim SS. Over- laxis (PrEP) to prevent HIV tunistic Infections, 23-26 coming Impediments to acquisition by high-risk Global Implementation Feb 2015, Seattle. MSM in England – results of Early Antiretroviral 18. Grant RM, Lama JR, Ander- of a static decision analyt- Therapy. N Engl J Med. son PL, et al. Preexposure ical model. Presented at 2015; 373:875-876. Chemoprophylaxis for Public Health England 13. WHO. Guideline on when HIV Prevention in Men Annual Conference, 15-16 to start antiretroviral Who Have Sex with September 2015, Coventry. therapy and on pre-expo- Men. N Engl J Med. 25. Saxton PJ, Dickson sure prophylaxis for HIV. 2010; 363:2587-2599. NP, Griffiths R, et al. 2015. http://www.who. 19. Grant RM, Anderson PL, Actual and undiagnosed int/hiv/pub/guidelines/ McMahan V et al. Uptake HIV prevalence in a earlyrelease-arv/en/ of pre-exposure prophy- community sample of 14. US Department of Health laxis, sexual practices, men who have sex with and Human Services and HIV incidence in men men in Auckland, New https://aidsinfo.nih. and transgender women Zealand. BMC Public gov/guidelines/html/1/ who have sex with men: Health. 2012;12:92. adult-and-adoles- a cohort study. Lancet 26. Saxton P, Dickson N, cent-arv-guidelines/10/ Infect Dis. 2014; 14:820-9. Hughes A, Ludlam A. initiating-art-in-treat- 20. McCormack S, Dunn DT, GAPSS/GOSS Research ment-naive-patients; Desai M, et al. Pre-ex- Brief: HIV testing among ASHM http://arv.ashm. posure prophylaxis to gay and bisexual men. org.au/clinical-guidance; prevent the acquisition of Auckland: The University British HIV Association HIV-1 infection (PROUD): of Auckland, 2015. https:// http://www.bhiva.org/ effectiveness results www.fmhs.auckland. HIV-1-treatment-guide- from the pilot phase of ac.nz/assets/fmhs/soph/

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 14 www.nzma.org.nz/journal EDITORIAL

sch/gmsh/docs/GAPSS%20 AIDS Organisations. HIV 39. Select Committee on HIV GOSS%20Research%20 testing among gay men and AIDS in the United Brief%20Prevalence%20 and other men who have Kingdom. No vaccine, no and%20Predictors%20 sex with men. Discussion cure: HIV and AIDS in the of%20HIV%20Testing%20 paper, 2014. http://www. United Kingdom. House of 2014%20final.pdf afao.org.au/library/topic/ Lords, 2011. http://www. 27. Gray RT, Holt, M, Lea T, et msm/HIV_Testing_DP_ publications.parliament. al. Estimated contribution ONLINE-July-2014.pdf uk/pa/ld201012/ldselect/ of undiagnosed HIV 33. Risher K, Mayer KH, ldaids/188/188.pdf infections among gay and Beyrer C. HIV treatment 40. New Zealand AIDS bisexual men to new HIV cascade in MSM, people Foundation. 2015 Annual infections in Australia. who inject drugs, and sex Report. https://www.nzaf. Presented at Australasian workers. Curr Opin HIV org.nz/assets/ee-uploads/ HIV&AIDS Conference, AIDS. 2015; 10:420-9. annual-reports-uploads/ 16-18 Sept 2015, Brisbane. 34. UNAIDS. 90-90-90 An NZAF_Full_Annual_ 28. Wilson DP, Hoare A, Regan ambitious treatment Report_2015.pdf DG, Law MG. Importance target to help end the AIDS 41. Covec. Cost Benefit of promoting HIV testing epidemic. Joint United Analysis of HIV Prevention for preventing secondary Nations Programme Programmes. Report transmissions: model- on HIV/AIDS (UNAIDS), prepared for New ling the Australian HIV 2014. http://www. Zealand AIDS Foundation. epidemic among men who unaids.org/en/resources/ Auckland: Covec, 2013. have sex with men. Sex documents/2014/90-90-90 42. Kippax S. Biomedical Health. 2009; 6: 19-33. 35. Blaxhult A, Bratt G. prevention: rhetoric and 29. Punyacharoensin N, Treatment as preven- reality. HIV Australia. Edmunds WJ, DeAngelis tion for HIV: a reality. 2015; 13 (2). https://www. D, et al. Modelling the HIV AIDS. 2015; 29:1721. afao.org.au/library/ epidemic among MSM 36. Phillips AN, Cambiano V, hiv-australia/volume-13/ in the United Kingdom: Miners A, et al. Potential vol-13-number-2-horizons/ quantifying the contribu- impact on HIV incidence biomedical-preven- tions to HIV transmission of higher HIV testing rates tion-rhetoric-and-reality#. to better inform preven- and earlier antiretroviral VlehKEbXtdA tion initiatives. AIDS. therapy initiation in MSM. 43. Juusola JL, Brandeau 2015; 29:339–349. AIDS. 2015; 29:1855-1862. ML. HIV Treatment and 30. Miller WC, Rosenberg 37. Shelton JD. Evidence- Prevention: A Simple NE, Rutstein SE, Powers based public health: not Model to Determine Opti- KA. Role of acute and only whether it works, mal Investment. Med Decis early HIV infection in but how it can be made to Making. 2015; online first. the sexual transmission work practicably at scale. 44. Cairns G, McCormack S, of HIV. Curr Opin HIV Global Health: Science Molina J-M. The European AIDS. 2010; 5:277-282. and Practice. 2014: 253-8. preexposure prophylaxis 31. Wilson DP. HIV Treatment 38. Bernard D, Kippax S, revolution. Curr Opin as Prevention: Natural Baxter D. Effective HIV AIDS. 2015; online Experiments Highlight partnership and adequate first. http://journals. Limits of Antiretrovi- investment underpin lww.com/co-hivandaids/ ral Treatment as HIV a successful response: Abstract/publishahead/ Prevention. PLoS Med. key factors in dealing The_European_preexpo- 2012; 9:e1001231. with HIV increases. Sex sure_prophylaxis_revo- 32. Australian Federation of Health. 2008; 5:193-201. lution_.99499.aspx

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 15 www.nzma.org.nz/journal EDITORIAL

Drug misuse in sport: a historical perspective David Gerrard

ABSTRACT This editorial draws comparisons between the recent revelations of drug misuse in Russian sport, and the State-sponsored programme of the former German Democratic Republic. While 50 years separates these two regimes, there are commonalities. The history of major incidents involving drug abuse by serious national players in sport suggests a 20-year cycle, with the GDR, China and now Russia employing similar strategies. These events underscore the value placed upon international sporting success by politicians.

ecent doping revelations, implicating all, history reflects examples of episodic, Russian athletes, have focused un- endemic drug misuse by major sporting Rfortunate, yet familiar, links to elite, nations in approximately 20-year cycles. In contemporary sport. In December 2014, a the decades between 1960–1980, the former German television documentary suggested German Democratic Republic (GDR) became that State-sponsored doping practices were responsible for a programme known offi- embedded in Russian sport, involving mem- cially as State-Plan 14.25, sanctioning the bers of the wider ‘athlete entourage’. Spurred delivery of various performance-enhancing by the clamour from an uneasy international drugs to young elite athletes. This was sporting community, the World Anti-Doping linked to an innovative national scheme of Agency (WADA) appointed an independent talent spotting that employed special ‘sport Commission of Inquiry, and one year later schools’ (Kinder und Jugendsportschulen) the outcomes of their investigation are now from which East German stars of the future in the public domain.1 The authors of this would emerge having undergone a battery report have declared a “…deep-rooted culture of physiological testing.2,3 The consequent, of cheating…” centred primarily on track and unparalleled Olympic success of female East field, but pointing the finger at the Mos- German athletes in swimming and athletics cow-accredited Laboratory. Prominent sports of the period was deemed a positive physicians, scientists, coaches, laboratory reflection of advanced talent-recognition, personnel and high-ranking officials have cutting-edge sport science, specialised been identified in a clandestine collaboration coaching and specialist sports medicine. reminiscent of the East German regime of the International observers of the time looked 1960s.2 Further, the Report describes State enviously at the GDR, but what was not complicity in an orchestrated programme known at the time was that the success of of sports drug misuse with clever cover-ups their athletes had been ‘underwritten’ by that hoodwinked the sporting public. These the use of performance-enhancing drugs. revelations now cast an ominous shadow As if we needed further reminding, sport over the unprecedented success of named was indeed a powerful political tool. Much Russian medalists at the 2012 London Olym- earlier, the so-called Nazi Olympics of 1936 pics. Downstream consequences for Russian illustrated the potency of major interna- participation at next year’s Olympic Games in tional sport, well ahead of contemporary Rio de Janeiro are currently under scrutiny by technology that provides instantaneous the International Olympic Committee and the updates and dissemination of results. International Athletics Federation. Concerned observers in 1936 were quick Pessimists would say that none of this to record “… from the proliferation of Nazi comes as too much of a surprise. After emblems around the Olympic stadium, to

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 16 www.nzma.org.nz/journal EDITORIAL

the mass Nazi salute from the huge crowds, swimming (FINA) and athletics (IAAF). A few to the esteemed place of Games Patronage years later, authorities uncovered wide- bestowed upon Hitler himself, the Berlin spread drug misuse implicating several Games remain tainted by a propaganda and coaches, many of whom had links to the political overtone that leaves little doubt former East German sports regime. Although as to the political purpose of this sporting not politically sanctioned, there were strong occasion.”4 The film of the Berlin Olympics comparisons made of many techniques remi- produced by Leni Riefenstahl, regarded by niscent of the former GDR. many as the most outstanding cinematic Comparisons between the actions of record of any Games, demonstrates the the Russian Ministry for Sport and the dramatic influence of the Nazi Party, prolific State- sanctioned policies of the former in the presence of the ubiquitous swastika.4,5 East Germany remain valid, despite the And so, until the late 1980s, the GDR intervening 50 years. The relationship rose from comparative athletic obscurity between physician and the athlete-patient to emerge as one of the most successful is articulated by the International Olympic sporting nations in history. International Committee and embodied in the Olympic prestige, closely aligned with sporting Movement Medical Code.9 This includes an success, became a mechanism to promote overarching statement of safety, ensuring socialist policy. GDR athletes of the period that, “… sport is practised without danger to were likened to “…missionaries validating the health of athletes and with respect for the superiority of socialism over capitalism fair play and sports ethics.” And further, the 3 ...” However, the ergogenic influence of relationship between athlete and healthcare agents, including anabolic androgenic provider is “… subject to mutual respect”. steroids, was not formally divulged until Typical of the attitude adopted by most the reunification of Germany, when official countries is the 2010 statement of the reports of the East German Ministry for State Medical Council of New Zealand.10 Entitled Security (Stasi) were released to the public. “Prescribing performance-enhancing medi- Bioethical considerations for the cines in sport”, this states: long-term health consequences of drug administration to young women escaped “Any doctor who knowingly prescribes, the judgement of those driving this regime. administers, traffics, supplies or Chronic anabolic androgenic steroid use, otherwise assists in the use of linked with increased risks of cardiovas- prohibited substances, for the delib- cular disease, liver problems, violent mood erate purpose of enhancing sports swings, virilisation in females and a clear performance and helping a sports link with certain forms of cancer5,6,7 were person to cheat, may be subject to highlighted by subsequent prosecutors. disciplinary proceedings and may It would seem that GDR physicians held be liable to a charge of professional little regard for these consequences. The misconduct.” world of clinical medicine and sport science The autonomy of physicians to practise still reels from the revelations. In this safely and in the best interests of their contemporary human experiment by the patients should never become influenced GDR, “…government policy, measured in by external, non-clinical agents. At the gold medals, gave scant regard to human highest level in New Zealand, those in posi- suffering and permanent disability.”2 tions of sports medical leadership remain In 1984, China—after an absence of 32 unconstrained to provide athletes with years—heralded its return to the interna- appropriate, quality care. While politics tional sporting fold with remarkable success and sport remain irrevocably linked, what at the Los Angeles Olympic Games. Chinese appears to have occurred in the context of athletes won 15 gold medals across a number Russian sport can never be condoned. In of sports, placing them fourth on the interna- contrast, the antics of our political leaders tional medal table.8 The rapid rise of Chinese basking in the Rugby World Cup success of female athletes, particularly in swimming the All Blacks are trivial by comparison and and athletics, drew closer inspection from raise nothing more than a wry, somewhat the international federations governing embarrassed grin.

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 17 www.nzma.org.nz/journal EDITORIAL

Competing interests: Dr. Gerrard reports he is currently the Chair of the World Anti-Doping Agency (WADA) Therapeutic Use Exemption Committeeand a member of the WADA Health Medicine and Research Committee, both voluntary positions. Author information: David Gerrard, Medicine, Dunedin School of Medicine, University of Otago Corresponding author: David Gerrard, Medicine, Dunedin School of Medicine, University of Otago [email protected] URL: www.nzma.org.nz/journal/read-the-journal/all-issues/2010-2019/2015/vol-128-no-1426-4- december-2015/6743

REFERENCES: 1. The Independent 5. Gerrard DF. Play- 2002; 24(1):178-181. Commission Review: Final ing Foreign Policy 9. Pipe A. Drugs, Sport, Report. World Anti-Doping Games: States, Drugs and the New Millen- Agency. Montreal. 2015 and other Olympian nium. Clin J of Sport Vices. Sport in Society. 2. Franke WW, Berendonk Med. 2000;10(1):7-8. B. Hormonal doping 2008;11(4) 459-466. 10. Fan Hong. Doping and and androgenization of 6. Ungerleider S. Faust’s Anti-Doping in Sport in athletes: a secret program Gold: Inside the China: an analysis of of the German Democratic East German doping recent and present atti- Republic Government. Clin machine. 2000 Rodales, tudes and actions. Sport in Chem 1997;43(7):1262-79. St Martin’s Press Society. 2006;9(2): 314-333. 3. Dickman S. East Germany: 7. Cole BC. The East Science in the disservice German Sports System: 11. Olympic Movement of the State. Science image and reality. PhD Medical Code. Internation- 1991; 254:26-27. dissertation 2000 Texas al Olympic Committee. Lausanne. 2009 4. Baker WJ. Notes, Docu- Tech University. ments and Queries: 8. Bowers LD. Abuse of 12. Medical Council of New New light on the Nazi Performance-Enhancing Zealand. Prescribing Olympics. J Sport Hist Drugs in Sport. Thera- performance-enhancing 1981;8:2, 118-120. peutic Drug Monitoring. medicines in sport. 2010

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 18 www.nzma.org.nz/journal ARTICLE

Thirty years of condom- based HIV prevention by gay men in New Zealand Anthony J Hughes, Peter J Saxton

ABSTRACT Three decades after the first government-funded HIV prevention campaign in 1985, gay and bisexual men (GBM) remain the population most at risk of infection in New Zealand. We review the major determinants of the elevated HIV risk for GBM, describe New Zealand’s prevention response over the first 30 years, and summarise the public health record. HIV incidence among GBM is driven by the heightened biological efficiency of HIV transmission during unprotected anal intercourse, dense sexual partnering networks, and endemic HIV prevalence. Responses in New Zealand have emphasised evidence-based primary prevention by condom use, which were implemented in communities and supported by comprehensive public health action. New Zealand has a good international HIV prevention record among GBM, however HIV diagnosis rates are now higher than they were during the epidemic nadir of the late 1990s. Lessons from the first three decades must underpin future HIV control efforts.

uman Immunodeficiency Virus three decades long and counting—condoms, (HIV) is a dangerous and tenacious testing and antiretrovirals will inevitably Hpathogen which is responsible for remain the core parts of our armamen- one of the worst pandemics in recorded hu- tarium against HIV because of gay men’s man history. Globally, 37 million people are unique nexus of vulnerabilities. As the last living with diagnosed HIV infection, 40 mil- overview of prevention in this population lion have died from AIDS-related illnesses, was in 1996,3 we review key principles and 2 million were newly infected in 2014.1 behind barrier-based HIV prevention for While it is true that anyone can be infected GBM in this country to underline the impor- by HIV, the full impact has been uneven, tance of existing programmes and prepare with some populations bearing a dispro- the ground for future initiatives. portionate burden. In particular, HIV has shown itself to be extremely well adapted to Biological dimensions sexual transmission between gay men. New Zealand’s response to HIV in the 30 of HIV’s threat years since the first government-funded Universal biological properties of HIV, prevention campaign in August, 1985, has regardless of its host population, help been effective, and we have a good inter- explain why the world is facing this unre- national record.2 Among gay, bisexual and lenting pandemic (Table 1). Ongoing HIV other men who have sex with men (GBM), spread is facilitated by its silent, asymp- this has been achieved by a condom-based tomatic nature for as much as ten years,4 primary prevention approach delivered its transmission potential during intimate through communities, supported by the sexual behaviours,5 and its elevated infec- creation of enabling environments, and tiousness soon after acquisition, when it is implemented in partnership with clinical frequently unrecognised.6 Subsequent to services. Advances in antiretroviral infection, a number of biological processes treatment have further extended HIV impacting on the immune system define prevention possibilities. Until a vaccine is HIV’s threat and distinguish it from most discovered—and this search is now also other viral infections. These include an

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 19 www.nzma.org.nz/journal ARTICLE

Table 1: Properties of HIV defining its biological threat

Property of HIV Implication Infection is frequently asymptomatic for Individuals unlikely to be aware they or their many years partners are infected, may unwittingly transmit virus.

Highly infectious early acute phase Hyperinfectious in first 3 months when infection usually unidentified by testing.

Sexual transmission only possible by Requires mucosal exposure at certain sites with specific acts/behaviour high density of receptor cells. Not spread casually so individuals can directly control exposure.

Uneven transmission probabilities Anal intercourse most efficient, oral sex very inefficient. Transmission most likely during most intimate penetrative behaviour.

Extremely high levels of genetic variation Harder for immune system to control, major and potential for recombination challenge for vaccine development, increased risk of developing drug resistance, and risk of superinfection with different strains.

Integrates directly into genetic material Infection is permanent and ineradicable. In latent of target cells state, the virus is undetectable by immune system.

Preferentially infects coordinator cell Central component of immune system is attacked type in immune system and permanently damaged.

Conceals presence so immune system Immune system unable to eliminate HIV infection does not recognise it naturally.

Increases acquisition and transmission Synergistically enhances other STI epidemics. risk of other STIs

Highest mortality of any viral infection If untreated HIV infection is almost invariably fatal. other than rabies extremely high genetic variability and odds of having HIV were 19.3 times higher capacity for recombination, leading to for GBM than the general population.10 extensive subtype diversity, and its ability Current global estimates have HIV inci- to integrate permanently into the host dence declining for almost all populations,1 genome.7 Over time, this results in profound with the exception of GBM in whom inci- immune system damage, culminating in dence is either static or rising.11 death from opportunistic infections and New Zealand GBM have also been seri- cancers in the absence of timely diagnosis ously affected by this epidemic. In studies of and treatment. sexual health clinic attenders, GBM are 40 times more likely to have HIV than hetero- Gay men’s heightened sexual men and women.12 Community studies in Auckland suggest that approxi- vulnerability to HIV mately 6% of sampled GBM are living with No group is more heavily impacted by the HIV, one-in-five of whom are unaware of sexual transmission of HIV than gay and their HIV-positive status.13 For those newly bisexual men, in whom the first cases were diagnosed, over a third are identified late identified. In every region where data are in the course of infection (CD4 count<350, available, GBM have a higher prevalence of or a median of 4 years post acquisition).14 HIV infection than the general population.8 Transmission of HIV within New Zealand This is seen across diverse settings, such is concentrated among GBM,15 who in 2014 as New York City, where the HIV case rate accounted for 80% of New Zealand’s local- for GBM is 140 times higher than among ly-transmitted epidemic, despite comprising heterosexual men,9 to sub-Saharan Africa, around 2.5% of the total population. where average HIV prevalence is 17.9% for Furthermore, the number of GBM living GBM compared to 5% for all adults,8 and in with diagnosed HIV is growing every year low and middle income countries where the and is estimated to have more than doubled

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 20 www.nzma.org.nz/journal ARTICLE

between 1999 and 2009.16 In contrast to this things being equal, HIV incidence for GBM hyperendemicity, locally-transmitted HIV in would be reduced by 80–98% if the rela- heterosexuals, people who inject drugs, and tively high per-contact HIV transmission sex workers in New Zealand has remained at probability of condomless RAI was mathe- very low levels over more than thirty years.17 matically modeled to be the same as that for 8 Three decades is sufficient time to be condomless vaginal intercourse. confident that this disproportionate impact Although early in the epidemic the is not an accident of history, but a real prevailing wisdom held that physical phenomenon requiring explanation. It is trauma and blood exposure were the now clear that the three central reasons are main causes of HIV transmission through the high biological efficiency of HIV trans- anal intercourse, the biological processes mission through receptive anal intercourse, behind the elevated risk of RAI are now the different sexual partnering dynamics more clearly understood. Like all viruses, seen in GBM communities, and the endemic HIV must first come into contact with the HIV prevalence that drives high numbers of specific target cells that it is genetically new infections. programmed to infect. For HIV, three main High efficiency of HIV transmission receptors exist (CD4+, CXCR4 and CCR5); by anal intercourse CD4+ and one other receptor is required for infection to occur.22 These receptors Contrary to widespread opinion,18 are found primarily on helper CD4+ cells most sexual behaviours are inefficient, in the immune system. The largest number implausible or impossible modes of HIV of immune cells (40–65% or more) occur transmission. The per-contact proba- in gut-associated lymphatic tissue (GALT) bility of infection for different sexual acts and the thin anorectal mucosal surface varies substantially.5 This is highest for is therefore highly susceptible to HIV receptive anal intercourse (RAI) without infection.23 High HIV viral loads are seen in a condom, which has an 18 times higher the gut mucosa following infection, more probability of HIV transmission compared than in either blood or semen.24 Significant with receptive vaginal intercourse without damage to the mucosal immune system a condom.19,20 At the other end of the in the gut persists, even in the presence of spectrum, oral sex has a very low to negli- long-term antiretroviral treatment (ART).25,26 gible probability of HIV transmission and kissing poses no risk.5 The per-partnership Despite the transmission risks of unpro- probability for unprotected RAI—the tected RAI, anal intercourse itself is an cumulative risk from repeated acts over important sexual activity for the majority time—is 40%.19 These probabilities are also of gay and bisexual men, just as non-re- heterogeneous.21 Amplifying factors for productive penile-vaginal intercourse is a anal intercourse include: sexual position highly valued sexual behaviour for most (receptive intercourse being substantially heterosexual men and women. In New riskier than insertive); the early acute Zealand, over 90% of GBM sampled in phase of HIV infection (when HIV viral community surveys had engaged in anal loads are highest and transmission proba- intercourse at least once in their lifetime.27 bility is up to 26 times greater than during Rates of recent practice with casual sex chronic infection);6 the presence of other partners have been rising over the last sexually transmitted infections (STIs) that decade, from 68% in 2002 to 76% in 2011.28 concentrate HIV in semen or immune cells Half (52%) reported that their first anal at the site of infection; and circumcision intercourse occurred by the age of 20; by status for the insertive partner.5 For prac- age 30 this was 84%.27 Sexual role versatility tical purposes, unprotected receptive and is common, with 17% being exclusively insertive anal intercourse are therefore receptive in anal intercourse, 30% being the only sexual acts conferring a mean- exclusively insertive, and 53% reporting ingful HIV risk for GBM, and likely account both insertive and receptive anal inter- for >99% of sexually-acquired HIV in this course with casual partners in the 6 months population. Furthermore, the excess HIV prior to survey.27 GBM’s unique ability for risk attributable to anal intercourse was sexual role reversal (being receptive then highlighted by a recent study: all other insertive, which is not possible in hetero-

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 21 www.nzma.org.nz/journal ARTICLE

sexual intercourse) is significant, as it places husband or friend they have sex with), and an individual at high-risk of acquisition over two-thirds report a casual partner in and then subsequently of HIV transmission, the preceding 6 months.27 Although many accelerating spread across the gay male GBM are in monogamous relationships, population. this overlap leads to approximately half Sexual networks in gay and of those in longer-term sexual relationships engaging in sexually non-exclusive bisexual male communities behaviour (much of this being mutually It takes at least two people to engage in agreed). From the early 2000s, internet anal intercourse. Consequently, under- dating, and more recently geo-location apps, standing the acquisition and transmission of have transformed the sexual marketplace STIs like HIV requires the study of partner- for everyone, but especially for ‘invisible’ ships, not just individual acts. Moreover, minority populations such as GBM, with the spread of infection beyond a few 53.4% reporting an active profile at the isolated cases is shaped by the aggregated time of survey in 2014, and 63.1% having and dynamic pattern of these partner- ever acquired a sexual partner online.32 ships across a community, which connects These apps have enlarged the pool of infected with susceptible individuals, and potential sexual contacts, and improved defines the potential for ongoing chains of the efficiency of partner acquisition, both transmission over time. It is this density of in terms of its immediacy and the ability to sexual connectivity that influences a given match sexual preferences.33 Furthermore, community’s overall incidence and prev- GBM surveyed on dating sites report lower alence once HIV has been introduced.29 testing and condom use and less favourable Likewise, at a personal level, this density attitudes to safe sex.34 While nationally situates someone close to or further away representative surveys overseas confirm from HIV transmission pathways, and in that heterosexuals also report many of these doing so determines the probability that the sexual behaviours, GBM populations are on next sexual partner will have undiagnosed average far more likely to do so.35 infection. This macro phenomenon is often referred to as a sexual network (its social Combined factors create equivalent popularised in the concept of ‘six conditions for explosive HIV degrees of separation’).30 epidemics Community studies in New Zealand over In the absence of effective interven- thirty years indicate that the sexual network tions, viral properties (the high HIV of gay and bisexual men differs markedly transmission efficiency of unprotected from that of the heterosexual population. RAI, the high infectiousness of HIV in the The distribution of sexual partner numbers early acute phase, asymptomatic infection, is right-skewed with a long tail, with a and permanent duration of infectivity in significant proportion reporting high absence of cure) combine with behavioural partner numbers (around 10% of sampled properties (the frequency of anal inter- GBM report more than 20 partners in the course, sexual role versatility, and dense last 6 months).27 Rapid partner turnover, sexual networks) to produce explosive short-gap lengths between partners, and HIV epidemics in GBM populations. As sexual mixing between individuals in epidemics mature over time, the high this ‘core group’ can create reservoirs of underlying prevalence of HIV infection in infection: 19.6% of this group reported an GBM communities propels high ongoing STI in the last year, compared to 4.6% of incidence. Furthermore, in many soci- those reporting one partner.31 Incursion eties, social, cultural and health system of HIV and STIs into the core group will be factors, such as homophobia and hetero- even greater if condom use is inconsistent, sexism, continue to hinder interventions affecting risk not only for themselves, but among GBM—such as the provision of also for GBM with more moderate rates of relevant safe sex advice and delivery partner change but who are sexually linked of timely HIV and STI screening.36 Even to them. Around half of sampled GBM when these services are provided, the report a current regular sexual partner sequelae of living as a minority—such as at the time of survey (either a boyfriend/ social isolation, poorer mental health and

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 22 www.nzma.org.nz/journal ARTICLE

Table 2: Twenty reasons to establish and maintain condom use for anal sex by gay and bisexual men

Effective Condoms prevent HIV transmission at both the individual and community level extremely successfully.

Simple Straightforward to deploy: acquire the condom, open the packet, put it on and add lubricant.

Verifiable Both partners can tell if they are being used, and effectiveness does not depend on accurate communication.

Safe No harmful health side effects or capacity to promote genetic resistance to HIV or STIs.

Controllable Easily manageable by couples on their own without any professional assistance or follow-up.

Sustainable Gay men can maintain consistent condom use for many years, up to three decades so far in New Zealand.

Inexpensive Affordable for individuals and governments, and usually provided free of charge in community venues.

Marketable Tangible product that can be easily visualised, promoted and distributed.

Empowering Enables gay men to have the sex life they want and minimises the need to alter sexual repertoire.

Acceptable Most New Zealand gay and bisexual men are using them, especially with casual partners.

Accessible Condoms can be made readily available to everyone without prescription or other restrictions.

Comprehensive An impermeable physical barrier to HIV that also offers substantial protection against other STIs during anal sex.

Ethical Use demonstrates mutual care for sexual partners and commitment to gay community health.

Scaleable Can be far more easily rolled out as a population-wide prevention programme than any other available option.

Universal Effectiveness does not depend on knowing personal HIV and STI status, or require disclosure to sexual partners.

Reliable Quality control is excellent and they fail extremely rarely when stored and used correctly.

Established Condom use has been actively promoted to gay men in New Zealand since 1987 and remains widely supported here.

Complementary Use does not in any way limit the effectiveness of clinic-based prevention programmes that rely on antiretroviral therapy.

Timely Condoms are used only during anal intercourse when HIV and STI transmission risk is highest.

Convenient Can easily be taken across borders and carried in pockets. substance use—can impact on gay men’s with over 50 partners in the last 2 years, ability to maintain HIV risk reduction HIV prevalence was 71%.38 In Scandinavia, practices. the estimated reproductive number of The speed and scale of HIV’s spread in HIV among GBM was 15 secondary cases GBM populations given these conditions is during 1981–2 prior to control efforts.39 well-documented. Retrospective analysis Phylogenetic analysis of HIV diagnoses of a San Francisco GBM cohort identified among GBM in the UK revealed that 29% that 28% were infected by 1981, before of HIV infections occurred within large AIDS was described, and incidence in 1982 clusters and of these, 20% occurred within 6 alone was an extraordinary 20%.37 Of those months of the index case.40 Contact tracing

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 23 www.nzma.org.nz/journal ARTICLE

of a young, newly-infected gay male in the key drivers of HIV spread among gay Wales uncovered a sexual network of 123 men previously described (Table 2). As the individuals and 15 new undiagnosed HIV recent Canadian consensus statement on cases.41 US projections suggest that even HIV transmission notes: relatively small annual incidence rates “[c]ondoms are a cornerstone of HIV of 2.4% can result in 40% of GBM being prevention. Latex and polyurethane infected by age 40.42 Similar statistics are condoms act as an impermeable currently being observed in nascent gay physical barrier through which HIV communities in Thailand,43 China,44 and cannot pass. When used correctly the Philippines,45 illustrating that the same and no breakage occurs, condoms trajectory of HIV and STI transmission could are 100% effective at stopping the be repeated now if the lessons of the last 30 transmission of HIV because they years are forgotten. prevent contact between HIV-containing body fluids and the target The prevention cells of an HIV-negative individual.”48 response in The centrality of condom use has been reiterated in a 2015 UNFPA, WHO and New Zealand UNAIDS position statement, highlighting While GBM remain disproportion- that condoms have averted approximately ately affected by HIV in New Zealand, 50 million new HIV infections globally since we have forestalled the worst scenarios the start of the epidemic.49 described above. Early, comprehensive Health promotion and sustained interventions based on Over the following decade, substantial condom promotion have been central to effort was put into developing a health New Zealand’s favourable record and are promotion-based framework for condom summarised below. uptake, with the aim of removing obstacles The early response to condom use by gay men.50 This utilised By the middle of 1987, the scientific the five-sector action framework of the evidence indicated that only anal inter- Ottawa Charter: the creation of supportive course—not other sexual activities or social environments; healthy public policy; partner numbers, per se—was the primary developing personal skills; reorienting cause of HIV infection.38 Evidence was health services; and promoting community also accumulating that condoms, if used action.51 Examples included countering correctly and consistently, worked very well anti-gay prejudice, reforming the Human to prevent the sexual transmission of HIV.46 Rights Act 1993, delivering public safe sex These twin developments enabled the New campaigns, providing community-based Zealand AIDS Foundation (NZAF) to focus on HIV testing and condom distribution promoting condom use for anal intercourse initiatives, and establishing the HERO as its central prevention strategy from that gay community development project. The point. This had several profound implica- subsequent success of these initiatives tions for control. First, it offered gay men demonstrated the fundamental impor- agency over their risk of acquisition and tance of aligning legal, social, cultural, transmission. Second, it required minimal health and peer incentives towards the alterations to gay men’s sexual repertoire, public health intervention of condom use making the intervention acceptable and by gay men. Health promotion advocacy sustainable. Third, it provided a single and like this has continued to the present day, easily communicated target for communi- with legal decisions such as NZ Police vs ty-led public health efforts. Fourth, it was Dalley illustrating the importance of the applicable to all GBM having anal inter- law underscoring rather than undermining 52 course, regardless of HIV or relationship public health goals. status, simplifying implementation.47 Peer-led and community-based Since then, condoms have proven to responses be an extremely effective sexual health Confronting HIV successfully also neces- intervention because they respond to all sitated a new approach to public health.

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 24 www.nzma.org.nz/journal ARTICLE

Figure 1: Strategic approach guiding New Zealand’s condom-focused HIV prevention response among gay and bisexual men in the last 30 years

Lacking the usual coercive levers (raising community-based initiatives, simultane- prices, regulating supply or imposing ously enabling a professionalised response legal sanctions), the principal tool for that retained community involvement and HIV prevention has been persuading gay accountabilities. and bisexual men to voluntarily adopt Increasingly comprehensive risk-reduction behaviours. Although heterosexual allies were, and continue to be vital approaches Taking a broad synopsis of the historical supporters, it was clear when HIV first record, Figure 1 encapsulates the strategic appeared that calls to change behaviour as approach taken to HIV prevention in New fundamental as intimate sexual practices Zealand, the skills that were used, and had to be championed by gay men them- the health intervention disciplines that selves.47 Personal access to community contributed. Early responses mobilised HIV institutions and connections with awareness in gay communities, utilised community gatekeepers were required for scientific evidence to inform GBM about credible public health advocacy. Peer-based HIV and how to avoid it, and advocated for prevention became the principal delivery structural reforms to enable GBM to uptake model, mirroring the informal efforts of gay health-seeking behaviours. These agendas activists early on. were largely executed by a community Similarly, sex between men had been illegal development, a health education, and until 1986 and discrimination on the grounds a health promotion approach respec- of sexual orientation legal until 1993, leaving tively. Aspects of all this work continue many gay men with reasons to be distrustful to be necessary as new generations of of government institutions.50 Governments GBM become sexually active, especially themselves were often not comfortable as mainstream media interest in HIV and developing the sexualised HIV prevention safe sex has waned. Most recently, condom resources that were most effective. This promotion efforts have intensified using led to the decision that peer-led responses social marketing approaches to achieve should be community-based and held at more effective scale and frequency.53 ‘arms-length’ from officials in order to This has been a response to increasing foster trust and encourage engagement with diversification of social media, and compe- prevention services.3 Funding an inde- tition involving less effective alternative pendent NGO, such as NZAF, to deliver the approaches to HIV prevention (such as majority of HIV-prevention services for GBM expecting a sexual partner to know and has been an important feature of peer-led, disclose their HIV positive status).

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 25 www.nzma.org.nz/journal ARTICLE

The overall public health goal of these safe sex, but still a significant achievement. efforts over the last three decades has been There has even been support for condom to use strategy51,54 to establish a robust use as a public health approach among GBM social movement supporting condom use who report often not using them.55 Factors for anal intercourse. This work has direct associated with condom use presented parallels to the promotion of immunisation elsewhere in this issue include attitudes against other communicable diseases for to safe sex and exposure to condom social the general population. Shared problems marketing,59 emphasising the importance include the requirement to maintain high of ongoing public health promotion to intervention coverage as visible disease maximise adherence.57 burden and public concern declines, and Attitudes to HIV and safe sex the need to counter opposition (eg, ‘anti- Strengthening gay community norms to vaxers’ opposed to childhood immunisation support condom use has been a key objective and ‘barebackers’ opposed to condom use). in New Zealand since 1987. There is almost universal personal acceptance of condoms as Achievements an effective way to avoid HIV transmission, Controlling HIV and most respondents also believe that other gay and bisexual men support condom use.58 Epidemiological analyses indicate that Thus, there is strong evidence of a culture New Zealand’s public health record at of condom use among New Zealand gay limiting infection among GBM ranks among men.58 Against this is a growing minority of the best in the world. By 1997, new HIV diag- GBM who do not perceive HIV to be a threat noses among GBM in New Zealand had fallen because of new treatments, increasing from to the lowest number since the peak in 1989 20% in 2002 to 38% in 2014.32 Some GBM (to just over 20 local infections per year), also continue to report difficulties using a result that was maintained until 2001.17 condoms, approximately a third report Sharp rises in new diagnoses in this country sometimes feeling under pressure not to since 2000, four years after the provision use a condom, and 1 in 10 report that sex of new antiretroviral therapies for HIV and isn’t always as safe as they want it to be.32 coinciding with the emergence of internet In the absence of comprehensive condom dating, were universally recorded in GBM promotion for the general public in New communities in Western Europe, the UK, US Zealand, social marketing from NZAF such and Australia, implicating shared factors as LYC (Love Your Condom) has aimed to 15 not unique to New Zealand. A steady state increase peer expectations for condom in new HIV diagnoses in New Zealand from use, improve access to free condoms and 2005 contrasted with continuing increases in promote user-efficacy. Countervailing initia- many countries. tives to maintain condom use will be needed Condom uptake as the visible consequences of HIV infection Equally, efforts to raise and then maintain decline, as the efficacy of treatments to condom use for anal intercourse by GBM in reduce HIV transmission and acquisition risk New Zealand rate as one of this country’s is promoted, and as internet pornography outstanding public health successes. Retro- depicting unprotected sex becomes even spective research suggests that condom more widespread. use at first anal intercourse rose from 28% in 1985 to 83% in 2005.27 Behavioural Conclusion surveillance conducted in community Three decades into New Zealand’s HIV settings indicates that frequent condom use epidemic, gay and bisexual men remain the (“always or almost always”) with casual population at greatest risk of infection. As in male sexual partners is approximately 85% all countries, the primary biological driver of in the 6 months prior to survey and has HIV transmission among GBM is unprotected been maintained at that level since at least anal intercourse, spread being sustained by 2002,28 albeit with a small decline in 2014. dense sexual networks, coupled with endemic Condom use is consistently lower among HIV prevalence. Taken together, these factors regular sexual partners at around a third result in a high number of GBM circulating of GBM, reflecting the contextual nature of in the sexual network with undiagnosed

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 26 www.nzma.org.nz/journal ARTICLE

infection. In response, condoms—if used the need for additional interventions. New consistently and correctly—are extremely prevention options include HIV antiretro- effective in preventing HIV transmission, virals taken as treatment for HIV-positive because they provide an impermeable individuals, or as pre- or post-exposure barrier to HIV during anal intercourse and prophylaxis for HIV-negative individuals, their protection does not rely on up-to-date which can substantially reduce HIV trans- knowledge of actual infection status. New mission risks. HIV testing itself is evolving Zealand’s successful HIV prevention record to offer more rapid and convenient access. over many years demonstrates how valuable Nonetheless, the high transmission effi- condom-centred programmes are if delivered ciency of anal intercourse and the dense at sufficient intensity and at scale, with sexual networks evident among GBM have not changed, and HIV still has the same wide, cross-sectoral buy-in and supported by basic biological properties. Thirty years enabling environments. Condoms also offer later, and in the absence of a vaccine or a large number of practical advantages over cure, these consistent features impose other prevention options, including being the limits on the ability of testing-based only HIV intervention that simultaneously interventions alone to control HIV among limits STI spread. GBM. It is therefore essential that primary A clear understanding of HIV prevention prevention barrier methods remain at the implementation at the community coalface center of HIV prevention efforts for GBM, must underpin strategies for future HIV and condom use continues to be strongly control. HIV diagnoses among GBM in 2014 supported. Our collective ability to do were the highest ever recorded in New so will determine the success of the next Zealand, posing urgent questions about phase of epidemic management.

Competing interests: Tony Hughes organised the first government-funded HIV prevention campaign delivered by NZAF in 1985. Since then, he has held the roles of Biomedical Coordinator, Research Director and Scientific Director at NZAF. His primary focus has been on utilising scientific knowledge about the HIV epidemic in gay men to sharpen strategic approaches to prevention. Dr Peter Saxton was formerly Senior Researcher at NZAF from 1997 to 2010 undertaking HIV research, policy analysis and advocacy. Acknowledgements: We would like to thank the staff, executive directors, boards of trustees, patrons, volunteers, donors and supporters of NZAF over the last three decades. We thank our collaborating partner the AIDS Epidemiology Group at the University of Otago, in particular A/Prof Nigel Dickson. We thank the Ministry of Health for core funding and the NZAF Fellowship at the University of Auckland for funding Dr Peter Saxton. We especially wish to acknowledge the gay community response to HIV in New Zealand and the leadership, passion and commitment shown by innumerable individuals and friends over the years. We dedicate this paper to Vern Keller, NZAF Librarian from 1991 to 2015 who passed away earlier this year. Author information: Anthony J Hughes, Scientific Director, New Zealand AIDS Foundation, Auckland; Peter J W Saxton, Director, Gay Men’s Sexual Health research group, Department of Social and Com- munity Health, University of Auckland, Auckland Corresponding author: Tony Hughes, New Zealand AIDS Foundation, PO Box 6663, Wellesley St, Auckland. [email protected] URL: www.nzma.org.nz/journal/read-the-journal/all-issues/2010-2019/2015/vol-128-no-1426-4- december-2015/6744

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 27 www.nzma.org.nz/journal ARTICLE

REFERENCES: 1. UNAIDS. 2014 Global F, Beyrer C. Elevated risk cember-2015/6745 Statistics. [online] for HIV infection among 18. National AIDS Trust. HIV: http://www.unaids. men who have sex with Public knowledge and org/sites/default/files/ men in low- and middle-in- attitudes 2014. London: media_asset/20150714_ come countries 2000-2006: National AIDS Trust, 2014. FS_MDG6_Report_en.pdf a systematic review. 19. Baggaley RF, White RG, PLoS Med. 2007;4:e339. 2. Sharples KJ, Dickson NP, Boily MC. HIV transmis- Paul C, Skegg DC. HIV/ 11. Beyrer C, Sullivan P, sion risk through anal AIDS in New Zealand: Sanchez J, et al. The intercourse: systematic an epidemic in decline? increase in global HIV review, meta-analysis AIDS. 1996;10:1273-8. epidemics in MSM. and implications for HIV 3. Davis P. (ed) “Intimate AIDS. 2013;27:2665-78. prevention. Int J Epide- Details and Vital 12. McAllister SM, Dickson NP, miol. 2010;39:1048-63. Sharples K, et al. Unlinked Statistics: AIDS, Sexu- 20. Grulich AE, Zablots- anonymous HIV preva- ality and the Social ka I. Commentary: lence among New Zealand Order in New Zealand”. probability of HIV sexual health clinic Auckland: Auckland transmission through attenders: 2005-2006. Int J University Press, 1996. anal intercourse. Int J STD AIDS. 2008;19:752-7. 4. Fraser C, Riley S, Anderson Epidemiol. 2010;39:1064-5. 13. Saxton PJ, Dickson RM, Ferguson NM. Factors 21. Vittinghoff E, Douglas J, NP, Griffiths R, et al. that make an infectious Judson F, et al. Per-contact Actual and undiagnosed disease outbreak control- risk of human immunode- HIV prevalence in a lable. Proc Natl Acad ficiency virus transmission community sample of Sci. 2004;101:6146-51. between male sexual men who have sex with 5. Patel P, Borkowf CB, partners. Am J Epide- men in Auckland, New Brooks JT, et al. Estimating miol. 1999;150:306-11. Zealand. BMC Public per-act HIV transmission Health. 2012;12:92. 22. Haase AT. Perils at risk: a systematic review. mucosal front lines for 14. Dickson N, McAllister AIDS. 2014;28:1509-19. HIV and SIV and their S, Sharples K, Paul C. 6. Miller WC, Rosenberg hosts. Nat Rev Immu- Late presentation of HIV NE, Rutstein SE, Powers nol. 2005;5:783-92. infection among adults in KA. Role of acute and New Zealand: 2005-2010. 23. Shacklett BL, Anton early HIV infection in HIV Med. 2012;13:182-9. PA. HIV Infection and the sexual transmission Gut Mucosal Immune 15. Saxton PJ, Dickson NP, of HIV. Curr Opin HIV Function: Updates on McAllister SM, et al. AIDS. 2010; 5:277-282. Pathogenesis with Implica- Increase in HIV diagnoses 7. Bushman FD, Nabel GL, tions for Management and among men who have sex Swanstrom R. HIV: From Intervention. Curr Infect with men in New Zealand biology to prevention and Dis Rep. 2010;12:19-27. from a stable low period. treatment. New York: Cold Sex Health. 2011;8:311-8. 24. Zuckerman RA, Whitting- Spring Harbor Press, 2012. ton WL, Celum CL, et al. 16. Saxton PJ, Dickson NP, 8. Beyrer C, Baral SD, Higher concentration of McAllister SM, et al. HIV van Griensven F, et al. HIV RNA in rectal mucosa prevalence among men Global epidemiology of secretions than in blood who have sex with men in HIV infection in men and seminal plasma, New Zealand 1985-2009: who have sex with men. 25 years of public health among men who have sex Lancet. 2012;380:367-77. monitoring. Int J STD with men, independent of 9. Pathela P, Braunstein SL, AIDS. 2012;23:274-9. antiretroviral therapy. J Infect Dis. 2004;190:156-61. Schillinger JA et al. Men 17. Dickson NP, Lee SKB, who have sex with men Foster T, Saxton PJ. The 25. Yukl S, Wong JK. Blood have a 140-fold higher risk first 30 years of HIV in and guts and HIV: pref- for newly diagnosed HIV New Zealand: Review erential HIV persistence and syphilis compared of the epidemiology. in GI mucosa. J Infect with heterosexual men NZMJ 128:1426 http:// Dis. 2008;197:640-2. in New York City. J www.nzma.org.nz/ 26. Estes J, Baker JV, Brench- Acquir Immune Defic journal/read-the-journal/ ley JM, et al. Collagen Syndr. 2011; 58:408-16. all-issues/2010-2019/2015/ deposition limits immune 10. Baral S, Sifakis F, Cleghorn vol-128-no-1426-4-de- reconstitution in the gut. J

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 28 www.nzma.org.nz/journal ARTICLE

Infect Dis. 2008;198:456-64. 34. Saxton P, Dickson N, from HIV sequences for 27. Saxton P, Dickson N, Hughes A. Who is a nationwide epidemic. J Hughes A. GAPSS 2006: omitted from repeated Infect Dis. 2011;204:1463-9. Findings from the Gay offline HIV behavioural 41. Knapper CM, Roderick J, Auckland Periodic surveillance among Smith J, et al. Investigation Sex Survey. Auckland: MSM? Implications for of an HIV transmission New Zealand AIDS interpreting trends. AIDS cluster centred in South Foundation; 2006. Behav. 2013;17:3133-44. Wales. Sex Transm 28. Saxton PJ, Dickson NP, 35. Glick SN, Morris M, Infect. 2008;84:377-80. Foxman B, et al. A compar- Hughes AJ. Location-based 42. Stall R, Duran L, Wisniews- ison of sexual behavior HIV behavioural surveil- ki SR, et al. Running in patterns among men who lance among MSM in place: implications of HIV have sex with men and Auckland, New Zealand incidence estimates among heterosexual men and 2002-2011: condom use urban men who have sex women. J Acquir Immune stable and more HIV with men in the United Defic Syndr. 2012;60:83-90. testing. Sex Transm States and other indus- Infect. 2014;90:133-8. 36. Ludlam AH, Saxton PJ, trialized countries. AIDS Dickson NP, Hughes AJ. Behav. 2009;13:615-29. 29. Pourbohloul B, Brunham General practitioner RC. Network models 43. van Griensven F, Holtz awareness of sexual orien- and transmission of TH, Thienkrua W, et tation among a community sexually transmitted al. Temporal trends in and internet sample of gay diseases. Sex Transm HIV-1 incidence and risk and bisexual men in New Dis. 2004;31(6):388-90. behaviours in men who Zealand. J Prim Health have sex with men in 30. Schneeberger A, Mercer Care. 2015;7:204-12. Bangkok, Thailand, 2006- CH, Gregson SA, et al. 37. Hessol NA, Lifson AR, 13: an observational study. Scale-free networks and O’Malley PM, et al. Lancet HIV. 2015;2:e64-70. sexually transmitted Prevalence, incidence, diseases: a description 44. Mayer KH. Editorial and progression of human of observed patterns of commentary: The next immunodeficiency virus sexual contacts in Britain tsunami? HIV spread in infection in homosexual Asian men who have sex and Zimbabwe. Sex and bisexual men in with men. Clin Infect Transm Dis. 2004;31:380-7. hepatitis B vaccine trials, Dis. 2014;58:1760-2. 31. Dickson N, Ludlam A, 1978-1988. Am J Epide- Saxton P, Hughes A. miol. 1989;130:1167-75. 45. Ross AG, Ditangco RA, Beli- mac JG, et al. HIV epidemic Self-reported STIs and 38. Winkelstein W Jr, Lyman in men who have sex with sexual health checks in a DM, Padian N, et al. men in Philippines. Lancet cross-sectional study of Sexual practices and Infect Dis. 2013;13:472-3. gay and bisexual men in risk of infection by the New Zealand. Sex Transm human immunodeficiency 46. Chetwynd J, Chambers Infect. 2015;91:49-54. virus. The San Francisco A, Hughes AJ. Condom 32. Saxton P, Dickson N, Men’s Health Study. use in anal intercourse amongst people who Hughes A, Ludlam A. Gay JAMA. 1987;257:321-5. identify as homosexual, Auckland Periodic Sex 39. Amundsen EJ, Stigum H, heterosexual or bisexual. Survey (GAPSS) and Gay Røttingen JA, Aalen OO. N Z Med J. 1992;105:262-4. men’s Online Sex Survey Definition and estimation (GOSS): Basic Frequen- of an actual reproduction 47. Parkinson P, Hughes T. cy Tables 2002-2014. number describing The gay community and Auckland: University past infectious disease the response to AIDS of Auckland, 2014. transmission: application in New Zealand. N Z Med J. 1987;100:77-9. 33. Saxton PJ, Dickson to HIV epidemics among NP, Hughes AJ. Trends homosexual men in 48. Loutfy M, Tyndall M, in web-based HIV Denmark, Norway and Baril JG, et al.Canadian behavioural surveillance Sweden. Epidemiol Infect. consensus statement on among gay and bisexual 2004;132:1139-49. HIV and its transmission men in New Zealand: 40. Leigh Brown AJ, Lycett in the context of criminal complementing loca- SJ, Weinert L, et al. law. Can J Infect Dis Med tion-based surveillance. Transmission network Microbiol. 2014;25:135-40. AIDS Care. 2015;27:762-6. parameters estimated 49. UNFPA, WHO, UNAIDS.

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 29 www.nzma.org.nz/journal ARTICLE

Position statement 53. Rich J. HIV prevention 2002-2011. Research on condoms and the among gay and bisexual brief to the Ministry of prevention of HIV, other men. Paper presented Health. Dunedin: AIDS sexually transmitted at Three Decades of Epidemiology Group, infections and unintended HIV in New Zealand University of Otago, 2012. pregnancy. UNFPA, WHO, HIV Clinical Update, [online] https://www.fmhs. UNAIDS, 2015 [online] 8th May 2015. [online] auckland.ac.nz/assets/ http://www.unaids.org/ http://player.vimeo.com/ fmhs/soph/sch/gmsh/ en/resources/presscentre/ external/130713756. docs/gapss-goss-attitudes- featurestories/2015/ sd.mp4?s=eea703dd- research-brief.pdf b3ed5aa75178674d7d- july/20150702_ 58. Lachowsky NJ, Saxton PJ, ace69a&profile_id=112 condoms_prevention Hughes AJ, et al. Frequent 50. Lindberg W, McMorland 54. Freedman L. Strategy: condom use with casual J. “From grass roots to A history. USA: Oxford partners varies by sexual business suits”: the gay University Press, 2013. position among younger community response 55. Adams J, Neville S. gay and bisexual men in to AIDS in Davis P (ed) Resisting the ‘condom New Zealand: national “Intimate Details and every time for anal sex’ behavioural surveillance Vital Statistics: AIDS, health education message. 2006–2011. Sex Health. Sexuality and the Social Health Education Jour- Online first http://dx.doi. Order in New Zealand”. nal. 2012;71:386-394. org/10.1071/SH14220 Auckland: Auckland 56. Steiner MJ, Cates W Jr, 59. Saxton P, Dickson N, University Press, 1996. Warner L. The real prob- Hughes A, Ludlam A. 51. Nutbeam D, Blakey V. The lem with male condoms Infrequent condom use concept of health promo- is nonuse. Sex Transm with casual partners tion and AIDS prevention: Dis. 1999;26:459-62. among New Zealand gay A comprehensive and inte- 57. Ludlam A, Saxton P, and bisexual men. NZ grated basis for action in Dickson N, Hughes A. Med J. 2015; 128 128:1426 the 1990s. Health Promot Attitudes towards safe sex http://www.nzma.org.nz/ Int. 1990;5:233-242. among men who have sex journal/read-the-journal/ 52. New Zealand Police with men in New Zealand: all-issues/2010-2019/2015/ v. Dalley, [2005] 22 Findings from the GAPSS vol-128-no-1426-4-de- C.R.N.Z. 495. and GOSS surveys cember-2015/6746

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 30 www.nzma.org.nz/journal ARTICLE

The first 30 years of HIV in New Zealand: Review of the epidemiology Nigel Dickson, Bible Lee, Timothy Foster, Peter Saxton

ABSTRACT AIM: To summarise findings of the epidemiology of AIDS and HIV infection in New Zealand. METHOD: Key results from reports of AIDS and diagnosed HIV infection are presented. Where appropriate, data on HIV diagnoses are reported for the period 2010–2014 to indicate the current pattern of diagnoses. RESULTS: New Zealand has a well-described low prevalence epidemic of HIV infection, mostly concentrated in sub-populations of men who have sex with men (MSM), and heterosexual individuals from sub-Saharan Africa and South-East Asia. The former is largely due to transmission within New Zealand, whereas the latter mostly occurred overseas, although the difference has been less marked in recent years. The number of notified cases of AIDS peaked in the late 1980s, and dropped dramatically in the mid-1990s due to the introduction of effective antiretroviral treatments. Presently, most cases of AIDS are in people with previously undiagnosed HIV infection. In contrast, currently the annual number of diagnoses of HIV infection is higher than in the late 1990s, due to more occurring among MSM. Over the past 30 years, each sub-epidemic has demonstrated a distinct pattern, reflecting different determinants. HIV among people who inject drugs, sex workers, children and the general population has been restricted to very low levels. CONCLUSIONS: Control of HIV in New Zealand is favourable compared to many countries, however challenges remain, especially in prevention among MSM, and more timely diagnosis for all, especially those heterosexually infected. National monitoring of the clinical outcomes of people diagnosed with HIV would provide an indication of the provision of effective care and allow international benchmarking.

hat is now known as the acquired AIDS Epidemiology Group (AEG) based at immune deficiency syndrome the University of Otago, Dunedin. The AEG’s W(AIDS) was first recognised as surveillance has been centered on case a clinical entity in 1981, and the first case reports of AIDS and newly diagnosed HIV diagnosed in New Zealand in 1983.1,2 The infection, supplemented by HIV prevalence human immunodeficiency virus (HIV) was studies in sentinel populations. The AEG has identified as the causative agent in 1984,3 also been involved in surveys of behaviours and HIV antibody tests to detect infection known to drive the spread of HIV and became available in New Zealand in 1985.4 testing patterns. Understanding the patterns of the epidemic Collectively these three components are in the population is important to develop now known as Second Generation HIV appropriate preventive control and treatment Surveillance.6 While both diagnoses of AIDS services.5 To this end, AIDS was made a noti- and HIV infection are included in surveil- fiable condition in 1983, however, HIV was lance, since the introduction of effective not due to concerns this might discourage antiretroviral treatment (ART) in the testing. Coded information on new diagnoses mid-1990s, the information obtained from from the laboratories undertaking confir- AIDS notifications has been less valuable in matory testing for HIV antibodies has been understanding the epidemic of HIV infection available since this began. than previously. Epidemiological surveillance of AIDS The findings from the AEG’s surveillance and HIV was initially undertaken by the have been regularly reported in the news- Department of Health, and since 1989 by the letter AIDS—New Zealand, but as there has

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 31 www.nzma.org.nz/journal ARTICLE

been no recent published review of the New categorisation of the regions is based on Zealand epidemic, we have taken oppor- the areas covered by the Regional Health tunity of the 30th anniversary of HIV testing Authorities that existed when surveillance in New Zealand to review the current of HIV was intensified in the 1990s, with epidemiology. the population of the Northern Region The aims of this article are therefore to being mainly made up of people living in (a) summarise key findings of the current Auckland. From the beginning of 2002, the epidemiology based on reports of diagnosed laboratories performing HIV viral load (VL) AIDS and HIV in New Zealand, (b) discuss testing have provided the codes—derived how these, and findings from other sources, in the same ways as for AIDS notifications inform HIV care and prevention needs and HIV information—of people having among particular groups in New Zealand, their first VL test in their laboratory. If it and (c) consider important areas for future appears through linking of the code to the epidemiological surveillance. AEG’s HIV database that a person having a VL test had not had a positive HIV antibody Methods test in New Zealand, information is sought from the clinician who requested the VL An individual with HIV infection is test. This was established initially to gain defined as having AIDS when he or she first information on people being cared for in develops one of a number of specific condi- New Zealand with HIV infections diagnosed tions uncommon in people with normal overseas, without having had an antibody immunity. Clinicians diagnosing AIDS are test in this country. However, VL testing has required to make an unnamed coded noti- increasingly been used to confirm new HIV fication to the Medical Officer of Health, infections, so is now an important source of which are then forwarded to the AEG; the people being first diagnosed here. code is based on the individual’s initials, gender and date of birth. The information Since 2005, information on the initial CD4 required includes key demographic char- cell count has been requested on people acteristics, the AIDS-defining condition and newly diagnosed with HIV infection in New 8 likely means of infection. Zealand; initially, this was only among those whose diagnosis was confirmed Since antibody testing for HIV infection through WB testing, but subsequently first became available in New Zealand through VL testing, if the infection was first in 1985, the number of people newly diagnosed in New Zealand. The initial CD4 diagnosed with HIV on the basis of a confir- count gives an indication of the stage of the matory Western Blot (WB) antibody test disease at diagnosis, and when less than 350 has been available from the two labora- cells per cubic milliliter, it is considered a tories undertaking this testing, Auckland City Hospital Virology Laboratory and the late diagnosis. Institute of Environmental Science and To compare the recent epidemic among Research Limited, Porirua. These labora- men who have sex with men (MSM) in tories have provided the age, sex and likely New Zealand with that of other high- means of infection of these people when it income countries, the annual diagnosis was provided to them. rate of HIV infection of MSM in selected The information was sent initially to the countries were compared. These diag- Ministry of Health, and since 1989, to the nosis rates were derived annually for each AEG. As with AIDS, information is only country from 2004–2013 using the number supplied to the AEG by code and never of diagnoses among MSM as the denomi- identified by name. Since 1996, the AEG nator and the number of men aged 15–64 has undertaken enhanced surveillance of as the numerator. Details of the method are HIV, whereby further information is sought reported in the Appendix. from the clinician who requested the test.7 Information has been collected since The additional information includes the 1998 from paediatricians, via the New site of and reason for the test, the infected Zealand Paediatric Surveillance Unit, on person’s ethnic group, district of usual resi- babies born to women with HIV diagnoses dence, and likely country of infection. The at the time of delivery.9

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 32 www.nzma.org.nz/journal ARTICLE

Table 1: Gender, likely means of infection, age (at diagnosis), and ethnicity of people diagnosed with AIDS in 2010–2014 and <2010 and in total. 2010–2014 <2010 Total

No. % No. % No. % Total 105 100.0 1,048 100.0 1,153 100.0

Gender Male 85 81.0 932 88.9 1,017 88.2

Female 20 19.0 113 10.8 133 11.5

Transgender 0 0.0 3 0.3 3 0.3

Likely means of infection Homosexual contact (MSM) 57 54.3 718 68.5 775 67.2

Homosexual (MSM) or IDU 1 1.0 14 1.3 15 1.3

Heterosexual contact 33 31.4 206 19.7 239 20.7

Injecting drug use (IDU) 1 1.0 24 2.3 25 2.2

Transfusion or blood product recipient 0 0.0 21 2.0 21 1.8

Mother to child transmission 0 0.0 18 1.7 18 1.6

Other 0 0.0 5 0.5 5 0.4

Unknown 13 12.3 42 4.0 55 4.8

Age at diagnosis (years) <5 0 0.0 12 1.1 12 1.0

5–14 0 0.0 10 1.0 10 0.9

15–19 0 0.0 4 0.4 4 0.3

20–29 9 8.6 159 15.2 168 14.6

30–39 21 20.0 399 38.1 420 36.4

40–49 41 39.0 298 28.4 339 29.4

50–59 20 19.1 123 11.7 143 12.4

> 60 14 13.3 40 3.8 54 4.7

Unknown 0 0.0 3 0.3 3 0.3

Ethnicity European 54 51.4 722 68.9 776 67.3

Māori 22 20.9 116 11.1 138 12.0

Pacific Islander 5 4.8 35 3.3 40 3.5

African 5 4.8 77 7.3 82 7.0

Asian 13 12.3 64 6.1 77 6.7

Other 5 4.8 27 2.8 32 2.8

Unknown 1 1.0 7 0.7 8 0.7 MSM = Men who have sex with men.

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 33 www.nzma.org.nz/journal ARTICLE

Figure 1: Annual number of diagnoses of AIDS and deaths among people notified with AIDS

We present and describe key findings income countries, is contemporaneous with on the reports of AIDS and diagnosed HIV the introduction of effective antiretroviral infection. As this is a surveillance report, therapy (ART). statistical testing is not undertaken. Where As well as reducing the number of people appropriate we have combined data with HIV infection progressing to AIDS, reported over the five-year period 2010– treatments available in the mid-1990s 2014 to give an indication of the current resulted in a marked improvement of the pattern of diagnoses. survival of people meeting the criteria for AIDS. As an indication of this, of those diag- Results nosed with AIDS in New Zealand in 1990, AIDS less that 10% were still alive five years later, while this was the case for over 70% of Overall, there have been 1,153 people people diagnosed a decade later in 2000. notified with AIDS to the end of 2014 (Table 1). Just over two-thirds (67.2%) of notifica- Ideally, people are diagnosed with tions were among gay, bisexual and other HIV infection before developing serious MSM infected through homosexual contact, infections that classify them as having with men and women infected through AIDS. However in the period 2010–2014, heterosexual contact the second largest 74.3% (78/105) had been diagnosed with group (20.7%). The age at diagnosis ranged HIV infection at same time or less than from less than one to 78 years of age, with 3 months prior to developing AIDS-de- a median of 39 years; although for most of fining conditions. Among many of those, these people, infection would have occurred an earlier HIV diagnosis and prior ART at a younger age, as the median time from could have avoided progression to AIDS, so HIV infection to the development of AIDS is earlier HIV diagnosis could be expected to around ten years in untreated young adults, reduce the annual number of AIDS notifi- and shorter in older people.10 cations further. The annual number of diagnoses of AIDS, HIV infection and deaths of people notified with AIDS, Information obtained from the HIV are shown in Figure 1. AIDS diagnoses testing laboratories indicates that 4,168 peaked in 1989 and 1990 with 71 cases, people have been diagnosed with HIV in and deaths in 1992 with 66. The dramatic New Zealand to the end of 2014 (Table 2). drop in the number of diagnoses in the Of these, 3,452 were through positive WB mid-1990s, which was seen in other high- antibody tests, and 716 through having

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 34 www.nzma.org.nz/journal ARTICLE

Table 2: Likely means of infection of people diagnosed with HIV in 2010–2014, <2010 and total. These figures include people previously diagnosed overseas. 2010–2014 <2010 Total

No. % No. % No. % Total 883 100.0 3,286 100.0 4,168 100.0 Homosexual contact (MSM*) 543 61.4 1744 53.0 2,287 54.8

Homosexual contact (MSM*) or IDU 6 0.7 42 1.3 48 1.2

Heterosexual contact 209 23.7 897 27.3 1,106 26.5

Injecting drug use (IDU) 8 1.0 76 2.3 84 2.0

Blood product/transfusion recipient 0 0.0 62 1.9 62 1.5

Mother to child transmission 3 0.3 55 1.7 58 1.4

Other 7 0.8 32 1.0 39 1.0

Unknown/Not stated 107 12.1 378 11.5 484 11.6

*MSM – Men who have sex with men

Figure 2: Number of people diagnosed with HIV in New Zealand, by year of diagnosis and means of infection. These figures include people previously diagnosed overseas.

VL tests among people not known by the infected through homosexual contact—a AEG to have had a prior WB test. While small number of whom were also reported scrutiny of codes have been used to detect to have injected drugs (Table 2); the duplicate reports, these have not always proportion rises to 63.4% if limited to those been provided, especially in the early with a reported means of infection. years of testing, so some duplication cannot be ruled out. The next largest group were heterosexually infected men and women, 26.4% of The annual number of diagnoses by means of infection is shown in Figure 2. It all diagnosed, and 30.0% of those with a is important to appreciate that for many, reported means of infection. Notably few the year of diagnosis will not have been the people have been definitely or possibly same as when the infection occurred, so is infected through injecting drug use. In not an indication of true annual incidence. 2010–2014, the proportion of diagnoses As with AIDS, the majority (56.0%) of among MSM has increased to 70.7% of those those diagnosed with HIV were MSM—men with a known means of infection.

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 35 www.nzma.org.nz/journal ARTICLE

Figure 3: Place of infection of MSM first diagnosed with HIV in New Zealand yb year of diagnosis, 1996–2014. These figures exclude people previously diagnosed overseas

Figure 4: International comparison of MSM diagnosis rate per 100,000 men aged 15–64, 2004–2013

Gay, bisexual and MSM Although the median age for HIV diag- Overall, there have been 2,335 MSM diagnosis among MSM was 37 years, the range is nosed with HIV. After an initial rise in the wide, with the youngest being 16 and oldest, annual number in the late 1980s and early 78 years. As well as appreciating that these 1990s, the number dropped to a nadir in are the ages at diagnosis not infection, it the late 1990s, with a subsequent rise in the needs to be kept in mind that this will not early 2000s (Figure 2). While there was a reflect the current age profile of MSM living steady increase in the years 2001 to 2005, with HIV, which will be older in view of the since then the annual number has fluc- success of current treatments. tuated. The highest ever annual number The ethnic profile of MSM diagnosed in of MSM was diagnosed in 2014, and could the five-years 2010–2014 (Table 3) is broadly indicate an upward trend in incidence, but similar to that of the male population it is too soon to conclude this. The overall aged 15–64 in the 2013 census. The higher rise since the early 2000s has been greatest proportion of an “other” ethnicity among among those infected in New Zealand rather HIV diagnoses is a reflection of people from than overseas (Figure 3). overseas diagnosed here. The increase in

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 36 www.nzma.org.nz/journal ARTICLE

Table 3: Characteristics of men who have had sex with men (MSM) diagnosed with HIV in 2010–2014, <2010 and total. These figures include people previously diagnosed overseas. 2010–2014 <2010 Total

No. % No. % No. %

Total 549 100.0 1,786 100.0 2,335 100.0 Age at diagnosis 15–19 11 2.0 34 1.9 45 2.0

20–29 142 25.9 468 26.2 610 26.1

30–39 151 27.5 652 36.5 803 34.4

40–49 148 27.0 382 21.4 530 23.0

50–59 64 11.7 155 8.7 219 9.1

60 or more 33 5.9 45 2.5 78 3.3

Unknown 0 0.0 50 2.8 50 2.1

Likely place of infection* New Zealand 353 64.3 655 58.2 1,008 60.2

Overseas 177 32.2 414 36.8 591 35.3

Unknown 19 3.5 57 5.0 76 4.5

Ethnicity* European 348 63.4 838 74.3 1,186 70.7

Māori 50 9.0 121 10.7 171 10.2

Pacific 19 3.5 36 3.2 55 3.3

Asian 84 15.3 77 6.8 161 9.6

African 0 0.0 7 0.6 7 0.4

Other 46 8.4 45 4.0 91 5.4

Unknown 2 0.4 4 0.4 6 0.4

Usual residence* New Zealand 516 94.0 1,030 91.3 1,546 92.2

Northern 303 58.7 528 51.3 831 53.7

Midland 41 7.9 137 13.3 178 11.5

Central 104 20.2 177 17.1 281 18.2

Southern 68 13.2 188 18.3 256 16.6

Overseas 28 5.1 68 6.0 96 5.7

Unknown 5 0.9 30 2.7 35 2.1

Initial CD4 count** <200 83 15.1 72 14.2 155 14.7

200–350 76 13.8 58 11.5 134 12.7

>350 274 50.0 193 38.1 467 44.3

Unknown 116 21.1 183 36.2 299 28.3

* Information on likely place of infection, ethnicity and usual residence collected since 1996 ** Information on initial CD4 count collected since 2005

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 37 www.nzma.org.nz/journal ARTICLE

Figure 5: Place of infection of people first diagnosed in New Zealand with heterosexually acquired HIV. These figures exclude those previously diagnosed overseas, 1996–2014

recent years in the proportion of Asian in New Zealand with heterosexually people likely reflects the changing ethnic acquired HIV who were infected overseas, make-up of Auckland, where over 50% of rather than in this country (Figure 5). newly diagnosed MSM and the HIV epidemic Figure 5 also shows that there was a marked in MSM is concentrated (Table 3). rise in the annual number in this group in Overall, of MSM diagnosed in 2010–2014 the period 2003–2006, with a subsequent for whom an initial CD4 count was provided, drop over the ensuing five years. This 36.7% had an initial count of less than rise and fall was due to an increase, 350 cell per cubic mm, hence considered and subsequent drop, in people coming late diagnoses; however, when this was to New Zealand from countries where restricted to those who had not been previ- heterosexually acquired HIV was relatively ously diagnosed (many of those diagnosed common, particularly sub-Saharan Africa. overseas would have been on treatment) the While the number of heterosexually proportion increased to 42.0%. acquired infections that have occurred in New Zealand remains relatively low, overall International comparison of HIV there has been a slight rise since 1996. diagnosis rates among MSM Similar numbers of men and women HIV diagnosis rates among MSM from 2004– have been diagnosed with heterosexually 2013 for the included countries are shown acquired HIV infection (Table 4). While in Figure 4. The rate of diagnosis in New overall most infections were acquired Zealand is lower than many of the countries overseas, the proportion was lower in examined. The US has a much higher rate of 2010–2014 (63.1% of all men and women for HIV diagnosis among MSM compared with all whom a place of infection was reported), other countries, while the Scandinavian coun- than in 1996–2010, when the comparable tries of Norway, Sweden and Finland have the proportion was 82.4%, due mainly to a lowest. Overall HIV diagnoses among MSM drop in overseas acquired infections rather in the countries examined rose slightly from than a rise in local ones. In addition, in the 2004–2013. In New Zealand over this period, most recent five-year period, about half the diagnosis rate shows moderate fluctuation of the men (46.0%) and a quarter of the as the numbers are relatively small, and the women (28.6%) diagnosed with hetero- only country showing a sustained substantive sexually acquired HIV infection were of drop has been Switzerland. European ethnicity, a proportional increase Heterosexually infected men and from earlier years mainly due to a drop women in the number of people of non-European In contrast to the situation among MSM, ethnicity. In the period 2010–2014, 61.3% of each year more individuals are diagnosed the heterosexually infected men and 60.3%

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 38 www.nzma.org.nz/journal ARTICLE

Table 4: Characteristics of heterosexually infected men and women diagnosed with HIV in New Zealand in 2010– 2014, <2010 and in total. These figures include people previously diagnosed overseas. Men Women Total

2010–2014 <2010 2010–2014 <2010

No. % No. % No. % No. % No. %

Total 111 100.0 440 100.0 98 100.0 457 100.0 1,106 100.0 Age at diagnosis 15–19 1 1.0 1 0.2 3 3.1 12 2.6 17 1.5

20–29 18 16.2 78 17.7 29 29.6 185 40.5 310 28.0

30–39 35 31.5 187 43 33 33.7 177 38.7 432 39.1

40–49 27 24.3 122 27.7 21 21.4 59 12.9 229 20.7

50–59 19 17.1 37 8.0 7 7.1 19 4.2 82 7.4

60 or more 11 9.9 15 3.4 5 5.1% 5 1.1 36 3.3

Likely place of infection* New Zealand 31 27.9 52 12.8 39 39.8 83 20.9 205 20.3

Overseas 73 65.8 334 82.0 58 59.2 298 75.1 763 75.3

Unknown 7 6.3 21 5.2 1 1.0 16 4.0 45 4.4

Ethnicity* European 51 46.0 110 26.8 28 28.6 76 19.1 265 26.0

Māori 4 3.6 11 2.7 8 8.2 18 4.5 41 4.0

Pacific 6 5.4 10 2.4 8 8.2 20 5.0 44 4.3

Asian 24 21.6 70 17.1 21 21.4 73 18.3 188 18.4

African 20 18 189 46.1 28 28.6 198 49.6 435 43.0

Other 6 5.4 16 3.9 5 5.0 12 3.0 39 3.8

Unknown 0 0.0 4 1.0 0 0.0 2 0.5 6 0.5

Usual residence* New Zealand 69 62.2 280 68.3 60 61.2 273 68.4 682 67.0 Northern 7 10.1 28 10.0 9 15.0 31 11.4 75 11.0

Midland 4 6.0 16 5.7 1 1.7 5 1.8 26 3.8

Central 44 63.7 194 69.3 37 61.7 186 68.1 461 67.6

Southern 14 20.2 42 15.0 13 21.6 51 18.7 120 17.6

Overseas 22 19.8 67 16.3 24 24.5 72 18.1 185 18.2

Unknown 20 18.0 63 15.4 14 14.3 54 13.5 151 14.8

Initial CD4 count** <200 33 30.0 55 28.6 25 25.5 35 18.4 148 25.0

200-–350 16 14.0 32 16.7 16 16.3 36 18.9 100 17.0

>350 31 28.0 47 24.5 27 27.6 63 33.2 168 28.4

Unknown 31 28.0 58 30.2 30 30.6 56 29.5 175 29.6

* Information on likely place of infection, ethnicity and usual residence collected since 1996 ** Information on initial CD4 count collected since 2005

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 39 www.nzma.org.nz/journal ARTICLE

Table 5: Characteristics of heterosexually acquired men and women infected in New Zealand, diagnosed in 2010–2014, <2010 and in total. These figures include people previously diagnosed overseas

Men Women Total

2010–2014 < 2010 2010–2014 < 2010

No. % No. % No. % No. % No %

Total 31 100.0 52 100.0 39 100.0 90 100.0 212 100.0

Age at diagnosis 15–19 1 3.2 0 0.0 3 7.7 5 5.6 9 4.2

20–29 6 19.4 11 21.2 8 20.5 46 51.1 71 33.5

30–39 16 51.6 16 30.8 16 41.0 21 23.3 69 32.5

40–9 6 19.4 18 34.6 7 18.0 12 13.3 43 20.3

50–59 1 3.2 3 5.8 3 7.7 5 5.6 12 5.7

60 or more 1 3.2 4 7.6 2 5.1 1 1.1 8 3.8

Ethnicity* European 21 67.7 33 63.5 16 41.0 32 38.6 102 49.8

Māori 1 3.2 6 11.5 8 20.5 15 18.0 30 14.6

Pacific 3 9.7 3 5.8 6 15.4 9 10.8 21 10.2

Asian 2 6.5 5 9.6 4 10.3 12 14.5 23 11.2

African 3 9.7 5 9.6 5 12.8 13 15.7 26 12.7

Other 1 3.2 0 0.0 0 0.0 2 2.4 3 1.5 * Information on ethnicity collected since 1996

Figure 6: Annual number of children diagnosed with perinatally-acquired HIV by year and place of birth. These figures include children previously diagnosed overseas.

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 40 www.nzma.org.nz/journal ARTICLE

Table 6: Information on children (under the age of 15) diagnosed with HIV in 2010–2014,<2010 and in total. These figures include children previously diagnosed overseas 2010–2014 <2010 Total

No. % No. % No. %

Total 4 100.0 70 100.0 74 100.0 Means of Infection Blood product recipient 0 0.0 8 11.4 8 10.8

Transfusion recipient 0 0.0 0 0.0 0 0.0

Mother-to-child transmission 3 75.0 55 78.6 58 78.4

Other 0 0.0 1 1.4 1 1.3

Unknown/Not stated 1 25.0 6 8.6 7 9.5

Age at diagnosis <1 year old 0 0.0 12 17.0 12 16.2

1–4 year olds 2 50.0 25 36.0 27 36.5

5–9 year olds 1 25.0 21 30.0 22 29.7

10–14 year olds 1 25.0 12 17.0 13 17.6

Ethnicity European 1 25.0 4 6 5 6.8

Māori 0 0.0 8 11 8 10.8

Pacific 0 0.0 3 4 3 4.0

Asian 0 0.0 6 10 6 8.1

African 3 75.0 33 47 36 48.6

Other 0 0.0 1 1 1 1.4

Unknown 0 0.0 15 21 15 20.3

Likely place of infection New Zealand 3 75.0 24 34.2 27 36.5

Overseas 1 25.0 30 42.9 31 41.9

Unknown/Not stated 0 0.0 16 22.9 16 21.6

of the heterosexually infected women for individuals were of African ethnicity, and whom an initial CD4 was available were that among the 39 women, 20.5% were of diagnosed late, higher than the proportion Māori, and 15.4% of Pacific, ethnicity. among MSM (36.7%). Children In the five-year period 2010–2014, There have been 74 children under similar numbers of men (31) and women the age of 15 diagnosed with HIV in New (39) have been diagnosed with HIV Zealand (Table 6). In the early years, many infection, reportedly heterosexually-ac- were infected through the receipt of blood quired in New Zealand (Table 5). While products to treat coagulation disorders, but again there was a wide range of ages at diagnosis, for both men and women there have been no new diagnoses of cases the most common age group was 30–39 where infection had been transmitted in years. About two-thirds (67.7%) of the this way since 1997. Overall, most children men were of European ethnicity, but this have been infected through mother-to- was the case for less than half (41.0%) of child, or perinatal, transmission—that is the women. It is also notable that eight— being born to an HIV-infected mother, three men and five women—of the 70 many of whom have come from overseas.

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 41 www.nzma.org.nz/journal ARTICLE

Table 7: Births in New Zealand to known HIV-infected pregnant women in 2010–2015, 1998–2010 and in total. This information has been collected since 1998. 2010–2014 <2010 Total

No. % No. % No. %

Total 43 85 128 Region of birth Northern 25 58.1 42 49.4 67 52.3

Midland 4 9.3 8 9.4 12 9.4

Central 11 25.6 20 23.5 31 24.2

Southern 3 7.0 15 17.7 18 14.1

Timing of diagnosis Before pregnancy 39 90.7 63 74.1 102 79.7

During pregnancy 4 9.3 22 25.9 26 20.3

Ethnicity of mother European 8 18.6 13 15.3 21 16.4

Māori 2 4.7 10 11.8 12 9.4

Pacific 5 11.6 3 3.5 8 6.3

Asian 9 20.9 11 12.9 20 15.6

African 19 44.2 47 55.3 66 51.6

Other 0 0.0 1 1.2 1 0.7

Antiretroviral to mother Yes 42 97.7 82 96.5 124 96.9

No 1 2.3 2 2.3 3 2.3

Unknown 0 0.0 1 1.2 1 0.8

Delivery Vaginal 21 48.8 32 37.6 53 41.4

Caesarian 22 51.2 52 61.2 74 57.8

Unknown 0 0.0 1 1.2 1 0.8

Infant breast fed Yes 0 0.0 1 1.2 1 0.8

No 42 97.7 83 97.6 125 97.7

Unknown 1 2.3 1 1.2 2 1.5

Whereas most data are based on the year Unit shows that from 1998 to 2014, there of diagnosis, the year of birth of perinatally- have been 128 babies born to women with infected children gives an indication of the HIV that had been diagnosed before or actual year of infection (Figure 6). While there during their pregnancy (Table 7). None of has been no perinatally-infected children these children are known to have acquired diagnosed who were born since 2007, as HIV, although it is too soon to be absolutely diagnosis can be delayed for many years, certain about this for some of those born there may be children with undetected HIV in 2014. The data collected also show that infection currently living in New Zealand. virtually all infected pregnant women Data collected from paediatricians via received ART and avoided breast-feeding, the New Zealand Paediatric Surveillance measures known to greatly reduce the risk

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 42 www.nzma.org.nz/journal ARTICLE

of HIV perinatal transmission. In recent other reasons not diagnosed until many years, more deliveries have been performed years after infection. However, as it is likely vaginally than by caesarian section than that most people with HIV will eventually previously the case, in line with current become symptomatic and diagnosed, the understanding that for women with a general pattern and trends in diagnosis rate well-suppressed VL, vaginal delivery does are likely to indicate those of true incidence not carry a high risk. over time. While no surveillance system Number of individuals living with will capture all data, cooperation between patients, laboratories, diagnosing physicians, HIV in New Zealand and the AEG, has meant that the system has The number of individuals living with a high level of completeness, even though diagnosed HIV in New Zealand will HIV is not a notifiable condition. be less than the total ever found to be Clearly, New Zealand has an HIV epidemic infected because of deaths from AIDS and concentrated among MSM. While an non-AIDS-related causes, and an unknown international comparison shows that the number going overseas. Ministry of Health diagnosis rate of MSM is in the middle to data show there were 2,059 adults (1,699 low range found in high-income countries men, 360 women) and 23 children receiving with accessible information on this, care subsidised ARTs the end of June 2015.11 This is needed in interpreting these data. Not was 192 more adults than at end of June all infections would have occurred in the 2014, giving an estimate at the end of 2014 country of diagnosis, and this is most likely of 1,963 on funded treatment. to impact on small countries, such as New Currently, just over 90% of people with Zealand and Sweden. Diagnosis rates will be HIV under the care of the Auckland Infec- dependent on testing rates, as when these tious Disease Unit are on ART. It is known are low the impact on changes in incidence that some people will opt out of specialist will be delayed. Also the proportion of care, or not enter it for other reasons, MSM may not be the same in all countries. which we estimate to be 5% of those diag- Nevertheless, the New Zealand diagnosis 12 nosed, based on UK data. So, assuming the rate being lower than in many high-income 2,059 adults receiving subsidised therapy countries is consistent with the findings of a represent 85% of those with diagnosed study of MSM in Auckland in 2011, in which HIV infection in New Zealand, there were HIV prevalence was 6.5%, lower than most 2,309 adults in New Zealand living with comparative studies in the US, Australia, UK diagnosed HIV at the end of 2014. If 20% of and France.13 infected people are undiagnosed—based The pattern of an increase in diagnoses on a 2011 Auckland study which found 21% in the early 2000s, followed by a plateauing, of MSM with HIV infection had not been is also similar to the experience in many diagnosed12—the total number of adults high-income countries.14 The rise is generally with HIV in New Zealand is 2,886, a prev- ascribed to a relaxation of the behaviour alence of 64 per 100,000 total population. changes take up by many MSM when AIDS Presuming the same proportion of infected first appeared, occurring after HIV was men and women in the population are on perceived as less threatening with the treatment and undiagnosed, the number of availability of better treatment; although men with HIV in New Zealand will be 2,381, the rise could have also been contributed and women 505; giving prevalence esti- to by changes to sexual partner acquisition mates among men aged 15 and over of 136 facilitated by internet dating. The subse- per 100,000, and among such women of 29 quent leveling out of annual HIV diagnoses per 100,000 population. is consistent with the relatively stable rates of condom use since 2002 in New Zealand Discussion behavioural studies among MSM.15 Another While ideally epidemiological surveillance likely driver of an increasing incidence in provides timely and accurate knowledge the early 2000s, is the rise in prevalence of of the actual annual incidence of HIV, it HIV among MSM, which would be antici- is not possible to measure this directly, pated with decreasing mortality and ongoing as many people are asymptomatic or for incidence. Both factors are likely to be

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 43 www.nzma.org.nz/journal ARTICLE

important. Evidence of the former is the heterosexually acquired HIV that occurred resurgence of other STIs among MSM, partic- in New Zealand since 1996 when this ularly syphilis, that occurred in the early information was first collected. However, 2000s in many countries, including New the numbers have averaged less than 16 Zealand.16 The ongoing occurrence of other annually over the past five years, only STIs among MSM infected with HIV is also about a quarter of the annual number of indirect evidence of the lack of universal MSM infected in New Zealand. When the condom use, although these do not neces- respective sizes of the heterosexual and sarily always indicate HIV transmission risk, MSM populations are considered, the risk as they could have been acquired among among heterosexuals is very much less than seroconcordant men. that of MSM. This lower perceived, and It is too soon to know if the rise in the actual, risk is no doubt why more of those number of MSM diagnosed in 2014 is an heterosexually infected are diagnosed late. indication of rising incidence, however it While testing is less widespread among heterosexuals, and therefore the proportion seem prudent to review, and if possible undiagnosed may be higher than among strengthen, prevention strategies now, in MSM, there is no evidence that this number case this is so. is large, or that HIV is spreading widely The information on new diagnoses in without being recognised. If it were, the recent year shows a clear increase the number of pregnant women being diag- number of Asian men with HIV infection. nosed through antenatal testing would be This is likely to be a reflection of the very much higher than it is, as would be increased size of this population, indicating the number of blood donors diagnosed (all that specific needs for care and prevention of whom are tested). Nevertheless, HIV among this sector of the population should should be considered when a person has be considered. been at risk, or has an illness consistent The profile of men and women hetero- with HIV infection or an opportunistic sexually infected is clearly different infection. The information on those infected from the MSM. They are more likely to heterosexually in New Zealand shows a have been infected overseas than in New wide age range, and while all ethnic groups Zealand, particularly for those diagnosed are affected, there appears to be a dispro- before 2011. The rise in the number and portionate number of African, Māori and proportion of HIV diagnoses of people Pacific women being diagnosed. heterosexually infected in the period While it is tempting to attribute the drop 2002–2006 was a reflection of the increase in maternally-infected children to the in migrants to New Zealand from parts of progressive introduction of a universal offer the world where heterosexual infection of HIV testing during pregnancy in New was particularly prevalent, especially some Zealand since 2006, in fact the number of areas of sub-Saharan Africa. It resulted in pregnant women diagnosed has been very Africans being the second largest ethnic low. In the five-year period 2010–2014, there groups diagnosed with HIV. The peak in were only nine women reported to the AEG heterosexual diagnoses in 2006 is likely as diagnosed through antenatal testing, less related to the requirement introduced that anticipated from estimates based on the in October 2005 that all people seeking a number of children infected in New Zealand visa to remain in New Zealand for a year prior to its introduction. Nevertheless, or more required an HIV test as part of as the universal offer of HIV testing has an immigration medical, which applied successfully been introduced and accepted to people already in New Zealand seeking by the vast majority of pregnant women, it to remain here, as well as those seeking seems sensible to continue to include this in entry. The drop after 2006 will be due to less antenatal care, as the additional cost of HIV migration from those high prevalence coun- testing when added to other antenatal tests tries, and that if HIV is diagnosed overseas, is low, and the possibility of the incidence it usually precludes entry. increasing remains. There has been a slight upward trend Another factor that has resulted in few in the number of people diagnosed with children being infected is that pregnant

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 44 www.nzma.org.nz/journal ARTICLE

women with diagnosed HIV are being cared a reduced level of virus in blood and in for in a way that is successfully minimising semen, has resulted in treatment of people the risk of perinatal transmission. There with HIV being incorporated as one of the have been over 120 babies born to women strategies of prevention, and is referred to with diagnosed HIV in New Zealand since as ‘treatment as prevention’. There have monitoring of this started in 1998, none of been attempts to determine the potential whom have been infected with HIV, less than effectiveness of this at a national level by the one percent rate generally reported. estimating the cascade of care—being the The small number of people diagnosed proportion who are first diagnosed with with HIV reported to have been infected HIV, referred to specialist care, retained through sharing of equipment used to inject in care, on ART, and subsequently have drugs is consistent with prevalence studies a fully suppressed blood viral load. The in this population, which have shown initial determination of the cascade in it to be less than one percent of people the US was disappointing, suggesting that using the needle and syringe exchange in 2011 only around a quarter of people scheme.17 This is undoubtedly due to the with HIV were on ART without detectable early enactment of legislation enabling the circulating virus.18 In other countries, needle and syringe exchange scheme prior this was very much better, with an esti- to HIV being established in this sector of the mated 58%, 53% and 58% on ART with a community. Although not specially sought fully suppressed viral load in the UK, the 19 in the information obtained about people Netherlands and France, respectively. newly diagnosed, sex workers have not In New Zealand, an unpublished study by been identified as a major factor driving the AEG that sought this information from HIV infection in New Zealand. people diagnosed over a recent eight-year period, suggests that we were nearer these The actual number of people currently European figures. However, we were living with diagnosed HIV is not known. unable to emulate the study, as unlike in These we have estimated to be 2,381 men the UK, HIV surveillance is not linked to a and 505 women, based on the number of unique number that allows the necessary subsidised ART as monitored by PHARMAC information to be collected, resulting in a and reported by the Ministry of Health, significant proportion of people for whom which assume 85% of people diagnosed the outcome could not be determined. and still living in New Zealand are on treatment. While this might be underestimated, as the criteria for being on ART has Conclusion in recent years become less stringent, it also Ongoing HIV surveillance shows that New takes into account that some people will Zealand has a well-described, mature, low have opted out of specialist care. Another prevalence, HIV epidemic with infection assumption, that 20% of infected people are concentrated among MSM, and hetero- undiagnosed, is based on the 21% found an sexual individuals from sub-Saharan Africa Auckland-based 2011 study of MSM, which and South-East Asia. The former is largely would be expected to get lower over time, if the result of transmission occurring within the incidence does not change significantly New Zealand, whereas the latter infections and survival continues. As well as allowing were mostly, but not universally, acquired estimates to be made, the data suggests that overseas. Over 30 years, each sub-epi- the annual costs of ART alone for each indi- demic has demonstrated a distinct pattern vidual on treatment is around $NZ 14,000 reflecting different determinants. The per year,11 although this does not take into prevalence of HIV among people who inject account possible confidential reductions drugs, sex workers and children has been in the cost of certain ART pharmaceuticals restricted to very low levels. negotiated by PHARMAC. When added to the The control of HIV achieved in New personal impact and costs of other aspects Zealand is favourable compared to many of medical care, this emphasises the need to countries, however several challenges prevent new infections wherever possible. remain, especially in prevention among The appreciation that people on treatment MSM and more timely diagnosis for all are significantly less infectious, as they have those infected. HIV testing of those who

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 45 www.nzma.org.nz/journal ARTICLE

have been at risk needs to continue, as international benchmarking. However, an unnecessarily high proportion of HIV this would only be feasible if there were infection is still diagnosed late, and some a way that information from diagnosed not before progression to AIDS. These individuals could be accessed through the people are missing opportunities for timely health system, such as by using the National HIV treatment for personal wellbeing and Health Index number in conjunction with prevention of secondary transmission. notification of HIV. Behavioural surveillance Deaths from AIDS are now rare but still also needs to continue to monitor sexual occur, and conversely, the number living and HIV testing behaviour, and should, with diagnosed HIV is increasing markedly where feasible, be linked to HIV prevalence each year, with considerable implications studies to estimate rates of undiagnosed for care and treatment costs. People with infection. Routine or regular phylogenetic analysis of newly diagnosed cases, that HIV are particularly infectious soon after has not so far been routinely undertaken, acquiring infection, often before being could be used to identify clustering of new diagnosed even with regular testing, so infections and local HIV subtype diversity. behaviours aimed to reduce the risk of Surveillance of other sexually transmitted transmission need to be promoted strongly infections needs to be strengthened so that among all at risk and control cannot be it can provide information on the rates based on diagnosis and treatment alone. among MSM, as these are an indication Current epidemiological surveillance of HIV risk. Importantly, all monitoring needs to continue. The addition of national systems need to be acceptable to those in monitoring of the clinical outcomes of the general population and those in most people diagnosed with HIV would assess affected communities, and keep individual’s the provision of appropriate care and allow confidentiality paramount. Appendix Method used for international comparison of diagnosis rates among MSM New Zealand, Australia, Belgium, Canada, Finland, France, Germany, Ireland, the Netherlands, Norway, Portugal, Sweden, Switzerland, the UK and the US were selected for comparison over the time period 2004–2013 having data on new diagnoses among MSM for the entire period. Publicly available data on HIV diagnoses were collected from the public health agencies of the countries selected. New Zealand’s HIV data were directly available to the AIDS Epidemiology Group, Australia’s from the Kirby Institute, Belgium’s data from the Institut Scientifique de Santé Publique, Canada’s from the Public Health Agency of Canada, Finland’s from the Terveyden Ja Hyvinvoinnin Laitos, France’s data from the Institut de Veille Sanitaire,9 Germany’s data from the Robert Koch Institut, Ireland’s data from the Health Protection Surveillance Centre, The Netherlands’ data from Stichting HIV Monitoring, Norway’s data from the Norwegian Institute of Public Health, Portugal’s data from the Instituto Nacional de Saúse, Sweden’s data from the Folkhälsomyndigheten, Switzerland’s data from the Bundesamt für Gesundheit, the UK’s data from Public Health England and the US’s data from the Centers for Disease Control. Whole country data were only available from 2008–2012 for the US; however, as it is a major comparable country to New Zealand, the US data are displayed, though they are not included in the statistical analyses. Additionally, 2013 data from Portugal are affected by reporting delays, so the 2013 Portuguese data were not included in the statistical analyses. For all countries, unknown or unreported mode of transmission cases were proportionally reallocated to aid comparisons between countries (except the UK and France, who report data adjusted for unknown mode of transmission cases and, for France, reporting delays). This reassignment may lead to slight biases for certain countries; for example, most unknown Swedish cases are for overseas acquired infections that may be different to domestic infections. However, reassignment prevents countries with more complete mode of transmission information having artificially higher rates of HIV infection among MSM. Data for all countries are presented as diagnoses per 100,000 men aged 15–65, with population data drawn from official national governmental statistical offices. LOWESS non-parametric smoothing of the data for all the countries was undertaken to produce a smooth curve representing the estimated underlying average diagnosis rate for the period 2004–2013 to give an indication of the overall trend.20 This was then used to examine trends in certain individual countries that might be exceptions to this. The weighting algorithm for the LOWESS smoothing was contained within the lowes’ function from the statistical package of R, with a bandwidth of 2/3 of the data points.21 Countries were not weighted by population.

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 46 www.nzma.org.nz/journal ARTICLE

Competing interests: Nil Author information: Nigel Dickson, Department of Preventive and Social Medicine, University of Otago, Dunedin; Bible Lee, Preventive and Social Medicine, University of Otago, Dunedin; Timothy Foster, Department of Preventive and Social Medicine, University of Otago, Dunedin; Peter Saxton, Social and Community Health, University of Auckland, Auckland. Corresponding author: Nigel Dickson, Department of Preventive and Social Medicine, University of Otago, Dunedin [email protected] URL: www.nzma.org.nz/journal/read-the-journal/all-issues/2010-2019/2015/vol-128-no-1426-4- december-2015/6745

REFERENCES: 1. Centers for Disease 6. World Health Organi- AIDS in San Francisco. Control. Kaposi’s zation and Joint United Nature 1989;338:251-3 sarcoma and pneumo- Nations Programme on 11. Ministry of Health. cystis pneumonia among HIV/AIDS. Guidelines Country Progress Report homosexual men – New for Second Generation New Zealand (HIV/ York City and California. HIV Surveillance. World AIDS) 2014. Wellington, MMWR 1981;30:305-8 Health Organization New Zealand. Available and Joint United Nations 2. Romeril KR. Acute HTLV from: http://www.health. Programme on HIV/ III infection (letter). NZ govt.nz/publication/ AIDS, Geneva 2000. Med J 1985;98:401 country-progress-report- 7. Paul C, Wilson M, Dickson new-zealand- hiv-aids-2014 3. Gallo RC, Montagnier L. N, Sharples K, Skegg DC. 12. Curtis H, Yin Z, Clay AIDS in 1988. Scientific Enhanced surveillance K, Brown AE, Delpech American 1988;259:40-48 of HIV infections in New VC, Ong E. People with 4. Carlson RV, Skegg DC, Paul Zealand, 1996-1998. NZ diagnosed HIV infec- C, Spears GF. Occurrence Med J 2000;113:390-394 tion not attending for of AIDS in New Zealand: 8. Dickson NP, McAllister specialist clinical care: the first seven years. MRC S, Sharples K, Paul, C. UK national review. BMC AIDS Epidemiology Group. Late presentation of HIV Infect Dis 2015;15:315 NZ Med J 1991;101:131-134 infection among adults in 13. Saxton PJ, Dickson NP, 5. World Health Organi- New Zealand: 2005–2010. Griffiths R, Hughes AJ, zation. Consolidated HIV Med 2012;13:182-189 Rowden J. Actual and strategic information 9. Dickson N, Paul C, undiagnosed HIV prev- guidelines for HIV in the Wilkinson L, Voss L, alence in a community health sector. Switzerland: Rowley S. Estimates of sample of men who have World health Organi- HIV prevalence among sex with men in Auckland, zation. Available from: pregnant women in New New Zealand. BMC Public http://apps.who.int/iris/ Zealand. NZ Public Health Health 2012;12:92 bitstream/10665/164716 Reports 2002;9:17-19 14. Saxton P, Dickson NP, /1/9789241508759_eng. 10. Bacchetti P, Moss AR. McAllister S, Sharples K, pd f?ua=1&ua=1 Incubation period of Hughes A. Increase in

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 47 www.nzma.org.nz/journal ARTICLE

HIV diagnoses among AIDS, 2013;24:791-798 Pozniak A. Large dispar- homosexual men in 17. Henderson C, Brunton C, ities in HIV treatment New Zealand from a Lauzon C. Final Report cascades between stable low period. Sex eight European and of the National Needle Health 2011;8:311–318 high-income countries– Exchange Blood-borne analysis of break points. 15. Saxton P, Dickson N, Virus Seroprevalence Conference presentation Hughes A. Location-based Survey [BBVNEX2009] at HIV Drug Therapy HIV behavioural surveil- to the New Zealand Glasgow Congress, 2014. lance among MSM in Ministry of Health. 2011. Journal of the Interna- Auckland, New Zealand Unpublished report to 2002-2011: Condom use tional AIDS Society 2014, the Ministry of Health stable and more HIV 17(Suppl 3):19507 testing. Sex Transm 18. Gardner EM, McLees MP. 20. Cleveland WS. LOWESS: Infect 2014;90:133-138 Steiner JF, del Rio C, & A program for smoothing Burman WJ. The spectrum 16. Psutka R, Dickson N, Azari- scatterplots by robust of engagement in HIV care ah S, Couglan E, Kennedy locally weighted regres- and its relevance to test- J, Morgan J, Perkins N. sion. Am Stat 1981;35:54 and-treat strategies for Enhanced surveillance 21. R Core Team. R: A prevention of HIV infec- of infectious syphilis language and envi- in New Zealand Sexual tion. Clinical infectious ronment for statistical Health Clinics. Interna- diseases, 2011;52:793-800 computing. R Core Team. tional Journal of STD & 19. Raymond A, Hill A, Vienna, Austria 2012.

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 48 www.nzma.org.nz/journal ARTICLE

Infrequent condom use with casual partners among New Zealand gay and bisexual men Peter J Saxton, Nigel P Dickson, Anthony J Hughes, Adrian H Ludlam

ABSTRACT AIMS: To identify predictors of non-condom use among gay and bisexual men (GBM) in New Zealand with casual male partners. METHODS: We analysed anonymous self-completed data from GBM who participated in the community- based Gay Auckland Periodic Sex Survey (GAPSS) and Internet-based Gay Online Sex Survey (GOSS), undertaken in 2014. Infrequent condom use was defined as not using condoms “always” or “almost always” during anal intercourse in the prior six months. RESULTS: Of the 1,912 GBM reporting anal intercourse with a casual partner, 27.2% reported infrequent condom use. Being recruited from Internet dating sites, Pacific ethnicity, having over 20 recent male partners, infrequent condom use with a current regular partner, or being HIV-positive were independently predictive of infrequent condom use. Conversely, being older, having a tertiary degree, using a condom at first anal intercourse, being exclusively receptive with a casual partner/s, and seeing condoms promoted through multiple channels predicted frequent condom use. Attitudes to condoms and safe sex were strongly predictive of actual condom use. CONCLUSIONS: Social marketing should target the modifiable predictors of condom use, such as attitudes to safe sex. Interventions also need to engage successfully with GBM reporting non-modifiable traits such as HIV-positive GBM.

ay, bisexual and other men who have as opposed to regular sexual partnering is sex with men (GBM) are at greatest unknown, GBM having casual sex are likely Grisk of HIV infection in New Zealand, to change partners more frequently than accounting for 2,329 HIV diagnoses from those with regular partners, increasing 1985 to the end of 2014,1 and roughly 80% the probability of encountering a sexual of the locally transmitted HIV epidemic.2 partner with undiagnosed HIV in the highly Almost all HIV transmission between GBM infectious early acute phase of infection. occurs through anal intercourse without Individuals are also less likely to be aware a condom, hence health promotion efforts of their casual partners’ sexual and HIV have sought to maximise and sustain con- testing histories. Identifying predictors dom use in this population.3 Trends in con- of unprotected casual anal intercourse dom uptake among GBM in New Zealand therefore helps HIV prevention agencies, have largely been high and stable,4,5 how- such as the New Zealand AIDS Foundation, ever in 2011 this dropped slightly among target and place condom social marketing. casual partners, and in 2014 the highest Previous New Zealand research in 1996 ever annual number of locally acquired HIV found that GBM on lower incomes, who 1 infections among GBM was diagnosed. were not gay-community affiliated, or who Public health consequently needs a had fewer sexual partners were more likely better understanding of current patterns in to never use condoms with casual partners.6 non-condom use. Although the fraction of An analysis of our own behavioural surveil- HIV infection events attributable to casual lance data among younger GBM aged

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 49 www.nzma.org.nz/journal ARTICLE

under 30 between 2006–2011 found that the previous six months. Casual partners condom use was lower among Pacific GBM were defined as men they had had sex and those with less education. Alterna- with no more than three times over this tively, condom use was higher among GBM period, and regular partners men they recruited from community venues, who had had sex with four or more times. If partic- tested HIV-negative, had a modest number ipants had engaged in anal intercourse of recent sexual partners, did not have sex with a casual and/or a current regular with women, or who used condoms with partner they were asked the sexual position any regular male partners.7 (receptive, insertive), and for each position The aim of this study was to investigate the frequency of condom use on a five point scale (always, almost always, about half the factors predicting recent non-condom use time, very rarely, never). The questionnaire with casual sex partners using a large and contained socio-demographic items and diverse sample of GBM recruited from items about sexual partnering, HIV and STI community and Internet settings in 2014. testing, frequency of exposure to condom social marketing, and attitudes to HIV, Methods condoms and safe sex. Data collection Analysis We analysed data collected from the The main outcome was “infrequent” 2014 round of the Gay Auckland Periodic condom use (“never”, “very rarely” or Sex Survey (GAPSS) and Gay Online Sex “half the time”) for any anal intercourse Survey (GOSS), an established behavioural role; “frequent” use being at least “almost surveillance system consistent with WHO/ always” or “always” used a condom. The 8 UNAIDS Guidelines. GAPSS participants denominator is respondents reporting at were recruited in Auckland by trained least one episode of anal intercourse with a recruitment staff during one week in casual partner in the previous six months. February, 2014, from a gay community fair Chi-squared tests explored the association day and subsequently at all gay bars (four) of condom use with demographic character- and sex-on-site venues (five) in that city. istics, sexual partnering, health screening, Eligibility criteria were being male, aged at social marketing exposure and attitudes. least 16 years, having had sex with a man This informed the multivariate logistic in the past five years, and had not already regression models of factors independently participated in GAPSS or GOSS that year. associated with infrequent condom use. Questionnaires were voluntary, anonymous Due to potential collinearity between safe and self-completed on site. Secure return sex attitudes and social marketing on boxes ensured privacy. Following GAPSS, condom use, we developed three models: (1) the same questionnaire was used for the containing attitudes and socio-demographic Internet-based nationwide GOSS over the variables; (2) containing social marketing next month that accessed participants and socio-demographic variables; (3) through banners on New Zealand Internet containing both attitudes and social dating sites and apps. These included marketing variables. Statistical analyses NZDating.com, Manhunt, Grindr, Jack’D, and data management were carried out Hornet and Growlr. Detailed methods are using Stata v.12.1 on non-missing data. provided elsewhere.9 Ethics approval was Results received from the University of Auckland The 2014 surveys attracted 3,141 respon- Human Participant Ethics Committee dents, of whom 1,912 had engaged in anal (#010738) and surveys were funded by the intercourse with a casual partner in the Ministry of Health. previous six months and reported on their Questionnaire condom use. Of these, just under three- Participants were asked the number, type quarters (72.8%) reported frequent condom (casual or regular) and nature of current use, and just over a quarter (27.2%) infre- regular relationships (“Boyfriend/long-term quent condom use. The latter equated to lover/life partner/civil union partner/ 16.7% of all GAPSS/GOSS respondents. husband”, hereafter “BF”; or “fuckbuddy/ In univariate analyses, infrequent friend I have sex with”, hereafter “FB”) in condom use varied significantly by

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 50 www.nzma.org.nz/journal ARTICLE

Table 1: Infrequent condom use by socio-demographic characteristics

Socio-demographic Number Reported infrequent X2 characteristics condom use (n,%) p-value Total 1,912 518 27.2

Recruitment site Offline: community event 485 106 21.9 *

Offline: bars 51 6 11.8

Offline: sex-on-site venue 125 13 10.4

Online dating site 1,244 393 31.6

Age group 16–29 819 227 27.7 Ns

30–44 553 141 25.5

45+ 488 139 28.5

Ethnicity European 1,370 371 27.1 *

Māori 173 66 38.2

Pacific 56 21 37.5

Asian 198 40 20.2

Other 72 11 15.3

Highest education qualification Less than tertiary degree 998 332 33.3 *

Tertiary degree or higher 863 175 20.3

Free time spent with other gay men None 76 18 23.7 Ns

A little 655 178 27.2

Some 590 154 26.1

A lot 510 143 28.0

Sexual identity Gay or homosexual 1,535 432 28.1 Ns

Bisexual or other 365 83 22.7

* p<0.001. Ns=not statistically significant.

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 51 www.nzma.org.nz/journal ARTICLE

Table 2: Infrequent condom use by behaviours and HIV screening

Behaviours Number Reported infrequent X2 and screening condom use (n,%) p-value Condom used at first anal intercourse with a male No 730 277 38.0 *

Yes 1,132 232 20.5

Number of male sexual partners in last 6 months One 146 41 28.1 Ns

2–5 795 204 25.7

6–10 436 112 25.7

11–20 269 72 26.8

21–50 188 64 34.0

>50 52 20 38.5

Partnering and protective behaviours in last 6 months Casual only or no current regular partner 986 253 25.7 *

Current BF and no anal intercourse with him 61 14 23.0

Current BF and frequent condom use with him 109 5 4.6

Current BF and infrequent condom use with him 238 98 41.2

Current FB and no anal intercourse with him 73 15 20.6

Current FB and frequent condom use with him 226 6 2.7

Current FB and infrequent condom use with him 171 116 67.8

HIV testing history Last tested HIV-negative 1,329 334 25.1 *

Diagnosed HIV-positive 108 51 47.2

Untested or no result 412 122 29.6

STI diagnosed in last 12 months No 1,539 384 25.0 *

Yes 290 108 37.2

* p<0.001. Ns=not statistically significant. BF=boyfriend-type regular partner. FB=friend with benefit-type regular partner

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 52 www.nzma.org.nz/journal ARTICLE

Table 3: Infrequent condom use by condom social marketing exposure

Condom social Number Reported infrequent X2 marketing exposure condom use (n,%) p-value Frequency of seeing condom promotion in last 12 months Very frequently 829 187 22.6 *

Often 515 139 27.0

Occasionally 333 107 32.1

Rarely 165 60 36.4

Never 39 18 46.2

Number of places recalled seen condoms promoted in last 12 months† None 89 40 44.9 *

1 466 152 32.6

2 260 73 28.1

3 325 76 23.4

4 267 69 25.8

5 287 52 18.1

6 178 45 25.3

* p<0.001. † Options included “promos at gay events”, “billboards or bus-stop adverts”, “condom packs”, “promos online or on a mobile app”, “posters”, “material at saunas or cruise clubs”.

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 53 www.nzma.org.nz/journal ARTICLE

Table 4: Infrequent condom use by attitudes to condoms, HIV and safe sex

Attitude Number Reported infrequent X2 condom use (n,%) p-value “HIV/AIDS is a less serious threat than it used to be because of new treatments” Agree/strongly agree 632 217 34.3 † Disagree/strongly disagree 1,244 291 23.4 “Condoms are OK as part of sex” Agree/strongly agree 1,796 449 25.0 † Disagree/strongly disagree 86 63 73.3 “If he doesn’t want to use condoms I won’t bother using them” Agree/strongly agree 404 270 66.8 † Disagree/strongly disagree 1,470 238 16.2 “We all have a shared responsibility to protect other gay and bisexual men by using condoms for anal sex” Agree/strongly agree 1,766 434 24.6 † Disagree/strongly disagree 105 73 69.5 “I don’t like wearing condoms because they reduce sensitivity” Agree/strongly agree 793 345 43.5 † Disagree/strongly disagree 1,070 165 15.4 “It’s no-one else’s business whether or not I use condoms” Agree/strongly agree 602 270 44.9 † Disagree/strongly disagree 1,257 240 19.1 “I would sometimes rather risk HIV transmission than use a condom during anal sex” Agree/strongly agree 241 148 61.4 † Disagree/strongly disagree 1,612 360 22.3 “The sex I have is always as safe as I want it to be” Agree/strongly agree 1,594 375 23.5 † Disagree/strongly disagree 262 130 49.6 “I would never be willing to use condoms for anal sex” Agree/strongly agree 111 62 55.9 † Disagree/strongly disagree 1,743 445 25.5 “A man who knows he has HIV would tell me he was positive before we had sex” Agree/strongly agree 757 227 30.0 * Disagree/strongly disagree 1,092 281 25.7

*P<0.05. † p<0.001. recruitment site, ethnicity and education associated with infrequent condom use status (Table 1), condom use at first inter- after controlling for respondent socio-de- course, HIV testing and STI history (Table 2), mographic characteristics (Table 5; two condom social marketing exposure (Table 3) were omitted as they obviously indicated and attitudes to HIV, condoms and safe sex non-condom use: “I would never be willing (Table 4). to use condoms for anal sex” and “I would Three separate models were then sometimes rather risk HIV transmission developed to investigate the relationship than use a condom during anal sex”). The between condom social marketing most strongly predictive was agreement exposure, attitudes to condoms, and other that “if he doesn’t want to use condoms I potential independent predictors. In the won’t bother using them” (AOR 6.8, 95%CI first model, six attitude items remained 5.0–9.1).

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 54 www.nzma.org.nz/journal ARTICLE

Table 5: Attitudes independently associated with infrequent condom use with casual partners*†c

Attitude Adjusted odds p-value for ratio (95% CI) variable “Condoms are OK as part of sex” Agree/strongly agree (ref) 1

Disagree/strongly disagree 3.7 (2.0–7.0) <0.001

“If he doesn’t want to use condoms I won’t bother using them” Agree/strongly agree 6.8 (5.0–9.1) <0.001

Disagree/strongly disagree (ref) 1

“We all have a shared responsibility to protect other gay and bisexual men by using condoms for anal sex” Agree/strongly agree (ref) 1

Disagree/strongly disagree 4.2 (2.3–7.7) <0.001

“I don’t like wearing condoms because they reduce sensitivity” Agree/strongly agree 2.6 (2.0–3.4) <0.001

Disagree/strongly disagree (ref) 1

“It’s no-one else’s business whether or not I use condoms” Agree/strongly agree 2.0 (1.5–2.7) <0.001

Disagree/strongly disagree (ref) 1

“The sex I have is always as safe as I want it to be” Agree/strongly agree (ref) 1

Disagree/strongly disagree 3.4 (2.4–4.8) <0.001

* Two attitude statements were omitted from the model because they would obviously be correlated to condom use, including “I would never be willing to use condoms for anal sex” and “I would sometimes rather risk HIV transmission than use a condom during anal sex”. † Socio-demographic variables included in the model were recruitment site, age group, ethnic group, education and sexual identity.

In the second model, twelve non-attitude months, using condoms infrequently with variables were entered into a multivariate a current BF or FB, or being diagnosed logistic regression, including socio- HIV-positive. Conversely, being older, demographic (recruitment site, age group, having a tertiary degree, using a condom ethnicity, education, sexual identity), at first anal intercourse with a male, being behavioural (condom use at first anal exclusively receptive with a casual partner/s intercourse, number of partners, recent during anal intercourse, or seeing condoms partnering history, anal intercourse promoted in multiple ways were predictive modality), condom social marketing of frequent condom use with a casual exposure (frequency recalling condom social partner (Table 6). marketing, number of different condom Thirdly, when attitudes were introduced social marketing avenues recalled) and into model two, each of the attitudes remained HIV testing variables. The model found significantly independently associated with infrequent condom use with a casual condom use (data not shown). However, partner was independently predicted by the effect of some of the variables in Table 6 being recruited from Internet dating sites, diminished or disappeared, suggesting that being of Pacific ethnicity, having 20 or their predictive effect may be due to their more male sexual partners in the last six correlation with unfavourable attitudes.

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 55 www.nzma.org.nz/journal ARTICLE

Table 6: Socio-demographic, behavioural, HIV and condom social marketing factors independently associated with infrequent condom use with casual partners

Factor Adjusted odds p-value for ratio (95% CI) variable

Recruitment site 0.003

Offline: fair day (ref) 1 Offline: bars and sex-on-site venues 0.5 (0.3–0.97) Online dating site 1.5 (1.1–2.1) Age group 16–29 (ref) 1 0.042* 30–44 0.9 (0.6–1.2) 45+ 0.7 (0.5–0.99) Ethnic group 0.0285 European (ref) 1 Māori 1.4 (0.9–2.0) Pacific 2.2 (1.1–4.4) Asian 0.9 (0.6–1.5) Other 0.5 (0.2–1.03) Highest education <0.001 Up to tertiary degree (ref) 1 Tertiary degree or higher 0.5 (0.4–0.7) Sexual identity 0.096 Gay or homosexual (ref) 1 Bisexual or other 0.8 (0.5–1.1) Condom used at first anal intercourse with a male <0.001 No (ref) 1 Yes 0.4 (0.3–0.5) Number of male sexual partners in previous 6 months 0.013 Up to 20 (ref) 1 More than 20 1.6 (1.1–2.4) Modality of anal intercourse with casual partners 0.019 Both insertive and receptive (ref) 1 Receptive only 0.6 (0.5–0.9) Insertive only 0.8 (0.6–1.05) Partnering and protective behaviours in last 6 months <0.001 Casual only or no current regular partner (ref) 1 Current boyfriend and no anal intercourse with him or only frequent condom use 0.4 (0.2–0.7) Current boyfriend and infrequent condom use with him 2.5 (1.7–3.5) Current fuckbuddy and no anal intercourse with him or only frequent condom use 0.2 (0.1–0.3) Current fuckbuddy and infrequent condom use with him 4.9 (3.3–7.4) HIV testing history 0.001 Last tested HIV-negative (ref) 1 Diagnosed HIV-positive 3 (1.8–5.0) Untested or no result 1.1 (0.8–1.5) Frequency of seeing condom promotion in last 12 months 0.086* For each decline in frequency seeing condom promotion 1.1 (0.98–1.3) Number of places recall seen condoms promoted in last 12 months 0.008d For each increase in number of places seen condoms promoted 0.9 (0.8–0.97)

* P-value is for variable entered as ordinal categories.

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 56 www.nzma.org.nz/journal ARTICLE

difficult.6 However, our findings about Discussion predictors are broadly consistent with In this large and diverse sample of gay, recent research among younger GBM from bisexual and other men who have sex with the earlier GAPSS and GOSS surveys under- men recruited from community and Internet taken in 2006–2011, in particular the role dating sites in New Zealand, around a of Internet dating sites, HIV test status and quarter (27.2%) used condoms infrequently, patterns of condom use across both casual that is never, very rarely or at most half the and regular sexual partners.7 Unlike that time in the six months prior to survey. analysis, our sample was smaller, so we did Being recruited from Internet dating sites, not analyse condom use predictors by sexual having 20 or more male sexual partners position; however, an advantage was our in the last 6 months, using condoms infre- ability to explore the relationship with atti- quently with a current regular BF or FB tudes and social marketing in more detail. partner, being diagnosed HIV-positive Gay men in New Zealand appear to or being of Pacific ethnicity were inde- report higher rates of casual partner pendently predictive of infrequent condom condom use than GBM in Australia,11 the use with a casual partner. Conversely, US12 and the UK,13 where similar surveys being older, having a tertiary degree, using have been conducted, although differences a condom at first anal intercourse with a in sampling, measurement and reporting male, being exclusively receptive with a likewise complicate comparisons. A casual partner/s during anal intercourse, or possible explanation is that since 1987, New seeing condoms promoted in multiple ways Zealand has focussed on a single message: was independently predictive of frequent avoidance of anal intercourse without condom use with a casual partner. Atti- a condom, and has supported this with tudes to condoms were strongly predictive comprehensive health promotion and social of actual condom use, and their effect marketing responses to reduce barriers and remained strong after taking into account maximise condom uptake. This approach both socio-demographic and behavioural recognises that condoms not only reduce factors. There appeared to be a strong personal risk, but also help break chains connection between the extent of exposure of HIV transmission in sexual networks to condom social marketing, attitudes to and thus have a collective protective effect condoms, and actual condom use. by preventing secondary infections. Else- Strengths of this study include the broad, where, HIV prevention organisations non-clinic based recruitment approach, have promoted more diversified non-con- the large and diverse sample, the anon- dom-based risk reduction approaches ymous and self-completed participation over time, such as encouraging HIV testing that should minimise reporting bias about and disclosure of HIV test status (“sero- sensitive behaviours, the question speci- sorting”),14 or promoting non-condom use ficity providing information on frequency with certain partner types, so long as rules of condom use during anal intercourse around testing, communicating and sexual with casual and with regular partner types, monogamy are adhered to (“negotiated and the range of potential predictors in the safety”).15 questionnaire. While the latter approaches have the Limitations include the non-random appeal of endorsing greater variety of HIV sampling, a standard obstacle to prevention options, a risk is that multiple researching this population,10 thus the public health messages are more resource findings may not be generaliseable to all intensive to execute successfully over the GBM. The data are based on self-report of long term, and if they communicate mixed private activities. Questionnaire space was messages or contradict each other (eg, limited so we were unable to explore the “use condoms … don’t use condoms…”) potential effect of mental health, discrim- may inadvertently de-emphasise condom ination, alcohol and recreational drug use use, or imbue condom use with negative on condom use. connotations (eg, “don’t you know your HIV A different outcome measure makes status?”, “can’t you trust your partner?”). comparisons with early local research Further research on the diverging nature

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 57 www.nzma.org.nz/journal ARTICLE

of HIV public health responses for gay about behavioural risk compensation men in different countries would improve (condom rates dropping in response to our understanding of different condom pharmaceutical interventions) and imple- use patterns and comparative HIV mentation, including adequate scale, actual epidemiology. and opportunity costs, clinical capacity and 23 Despite this evidence of diversity in clinical retention. condom uptake between GBM, general HIV prevention has consequently become population probability studies have shown an intensely contested marketplace. We that gay and bisexual men are more likely recommend that the latter interventions than heterosexual men to use condoms for could build onto, but should categorically sexual intercourse, at least in countries not replace nor cannibalise, behaviour- where the HIV prevention responses have based condom promotion. We recognise been peer-led by GBM themselves.16 This that this is a difficult task for public health is not necessarily true for younger GBM, NGOs in a context of modest resources, and such as those at secondary school, who are growing pressure from some quarters for unlikely to have been offered relevant sex pharmaceutical interventions to replace education.17 Therefore, the majority of gay behaviour change. In the UK, the funding and bisexual men continue to demonstrate allocations are already stark: just £10 protective behaviour when supported to do million was allocated to behaviour-based so, and this needs to be remembered when prevention in high prevalence local author- seeking explanations for the overrepresen- ities in 2014/15, a reduction from £55 tation of GBM in HIV statistics. million in 2001/2, whilst spending on HIV Our results should be used to improve pharmaceuticals in 2014/15 was 55 times targeted HIV prevention with GBM commu- this amount.24 nities. An important general finding is Unsurprisingly, attitudes predicted that greater exposure to condom social condom use, but we highlight three obser- marketing is associated with more frequent vations. Firstly, the statements relate to condom use with casual partners. This different dimensions of using condoms—for argues for strategies to increase exposure example physical sensation (“I don’t like of condom promotion through multiple condoms because they reduce sensitivity”), channels that is a feature of New Zealand’s altruism/collectivism (“we all have a shared rejuvenated response to the HIV epidemic.18 responsibility…”), and personal resilience It does, however, pose a challenge to (“if he doesn’t want to use condoms…”). non-Government organisations (NGOs), This suggests it is an oversimplification such as NZAF, who are endeavouring to to assert that condom non-use is purely supplement condom-based behavioural due to GBM not liking them. Secondly, and HIV prevention with increased promotion encouragingly, attitudes are modifiable,25 of timely HIV testing, advice about the and in our data appeared to be influenced prevention and personal benefits of early by social marketing, and therefore poten- HIV treatment,19,20 and the potential benefits tially by community norms. More nuanced of HIV pre-exposure prophylaxis (PrEP) prevention responses that engaged GBM for high risk GBM under close specialist more compellingly about their sexual clinical monitoring,21,22 in order to reduce decision-making could therefore result in HIV incidence. In clinical trials, pharma- increased condom use. Thirdly, although ceutical-based prevention (“treatment some unfavourable attitudes were only held as prevention”) has reduced HIV trans- by a minority of respondents (less than five mission risk by 96% in HIV sero-discordant percent), others such as “if he doesn’t want heterosexual couples when used by the to use condoms…” were held by around a HIV-positive partner,19 and by up to 86% fifth of respondents and exerted a strong when used by GBM as pre-exposure prophy- impact on condom use. Understanding laxis by the HIV-negative partner.21,22 While better these more common attitudes, promising at the individual level, these trial and successfully challenging them, could results have yet to translate into wide scale potentially shift a high proportion of reductions in HIV incidence at the popu- the infrequent condom use that is being lation level among GBM, amid concerns reported by these GBM.

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 58 www.nzma.org.nz/journal ARTICLE

Condom use with casual partners was Respondents with previously diag- related to behaviours with other partners, nosed HIV infection reported less frequent supporting previous work indicating a condom use. Although some may have strong patterning or habitual factor to been serosorting (choosing other HIV-pos- condom use across partnerships for many itive casual partners), for others this GBM.26 This was reinforced by the associ- poses a risk of onward HIV transmission ation between early adoption of condoms especially as not all of these men were on and current use. Both findings argue for antiretroviral therapy. Furthermore, the continued promotion of condoms for anal risk of acquiring and transmitting other sex between men regardless of partner type, STIs during anal intercourse is high in the because stopping condom use in one circum- absence of condoms. HIV-positive GBM stance (for example with a regular partner) have been disproportionately represented may make it more difficult to continue in outbreaks of STIs such as syphilis and condom use in other scenarios (with LGV in New Zealand,27,28,29 and it is imper- sequential or concurrent casual partners). ative that clearer information about these Respondents with higher numbers of recent risks is communicated to HIV-positive GBM sex partners were more likely to report alongside improved linkage into support infrequent condom use, and it is important and screening. to engage these men effectively as they will Over and above this study’s findings, play a disproportionate role in facilitating HIV prevention for GBM must consider a or constraining HIV transmission clustering context of alarming rises in STIs. Among across GBM sexual networks. GBM in England for example, syphilis Above and beyond these factors, GBM diagnoses increased by 46% in the past recruited from Internet dating sites, year, gonorrhoea diagnoses by 32% and Pacific-identified respondents, younger chlamydia diagnoses by 26%. GBM now respondents and those with less than a account for 81% of syphilis and 52% of tertiary degree, were more likely to report gonorrhoea cases diagnosed in English infrequent condom use. Prevention inter- sexual health clinics.30 The emergence of ventions will need to ensure they engage antimicrobial resistance in gonorrhoea is a successfully with these groups. Internet serious concern. Reducing the incidence of dating sites and online geo-location apps condomless anal sex will help control these have had an especially profound impact on bacterial STIs, as well as HIV. GBM should HIV prevention for small communities, such be encouraged to screen for STIs regularly as gay men.5 Such apps expand the sexual which may also identify undiagnosed HIV marketplace and improve the efficiency of infections in a more timely way. finding a compatible partner with shared interests. However, these same features Finally, continued refinement and scale-up facilitate contact between men who wish of condom social marketing needs to be to engage in condomless sex. Furthermore, accompanied by better information about many apps do not seek meaningful part- why condom use is differentially important nerships with HIV prevention or sexual for GBM (elevated risks of anal intercourse, health agencies, unlike the cooperation dense sexual networks, higher HIV preva- historically shown by physical gay venues. lence) compared to their heterosexual peers. Local innovative HIV prevention responses In addition, health promotion for GBM must into the digital space, such as LYC (“Love continue the considerable progress made Your Condom”), have fostered closer rela- over three decades in New Zealand so that tionships (engaging dating apps, facebook, individuals are empowered to take up the google ads, websites),18 but several sites advice. This includes anti-discrimination and restrict the promotion of HIV prevention mental health promotion at high schools, material, raising questions about the public easier access to relevant information and health responsibilities of social media and health services, and sexual orientation commercial internet dating sites. diversity training for health professionals.

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 59 www.nzma.org.nz/journal ARTICLE

Competing interests: Nil Acknowledgements We would like to thank all participants, community venues and websites involved in the study. The GAPSS and GOSS projects are a collaboration between the Gay Men’s Sexual Health research group, the AIDS Epidemiology Group and the New Zealand AIDS Foundation. The Ministry of Health funded the 2014 study and the NZAF Fellowship assisted with analysis. Author information: Peter J W Saxton, Director, Gay Men’s Sexual Health research group, Department of Social and Community Health, University of Auckland, Auckland; Nigel P Dickson, Director, AIDS Epidemiology Group, Department of Preventive and Social Medicine, University of Otago, Dunedin; Anthony J Hughes, Scientific Director, New Zealand AIDS Foundation, Auckland; Adrian H Ludlam, Research Assistant, Gay Men’s Sexual Health research group, Department of Social and Community Health, University of Auckland, Auckland. Corresponding author: Dr Peter Saxton, Department of Social and Community Health, University of Auckland, Private Bag 92109, Auckland. [email protected] URL: www.nzma.org.nz/journal/read-the-journal/all-issues/2010-2019/2015/vol-128-no-1426-4- december-2015/6746

REFERENCES: 1. AIDS – New Zealand. 5. Saxton PJ, Dickson STI Surveillance; 2000. Issue 74. Dunedin: AIDS NP, Hughes AJ. Trends 9. Saxton P, Dickson N, Epidemiology Group, in web-based HIV Hughes A, Ludlam A. University of Otago, 2015. behavioural surveillance Gay Auckland Periodic among gay and bisexual 2. Saxton PJ, Dickson NP, Sex Survey (GAPSS) and men in New Zealand: McAllister SM, et al. Gay men’s Online Sex complementing loca- Increase in HIV diagnoses Survey (GOSS): Basic tion-based surveillance. among men who have sex frequency tables 2002- AIDS Care. 2015;27:762-6. with men in New Zealand 2014. Auckland: The from a stable low period. 6. Reid A, Hughes A, Worth University of Auckland, Sex Health. 2011;8:311-8. H, et al. Male Call Waea 2014. https://www.fmhs. Mai, Tane Ma Report Four: 3. Hughes A. Effective auckland.ac.nz/assets/ Casual sex between men. strategies for HIV preven- fmhs/soph/sch/gmsh/docs/ Auckland: New Zealand BFReport_34LoRes.pdf tion in gay men: The next AIDS Foundation, 1997. decade. Paper presented 10. Hughes A, Saxton P. 7. Lachowsky N, Saxton P, at Three Decades of Geographic micro-cluster- Hughes A, et al. Frequent HIV in New Zealand ing of homosexual men: condom use with casual HIV Clinical Update, 8th Implications for research partners varies by May 2015. https://www. and social policy. Soc sexual position among eiseverywhere.com/ Pol J New Zealand. younger gay and bisexual 2006; 28:158-178. file_uploads/5b565af98d23 men in New Zealand: 82f05749dfc076065f47_ national behavioural 11. de Wit J, Mao L, Adam TonyHughes.pdf surveillance 2006-2011. P, Treloar C. HIV/AIDS, 4. Saxton PJ, Dickson NP, Sex Health, 2015; Online hepatitis and sexually Hughes AJ. Location-based first http://dx.doi. transmissible infections in HIV behavioural surveil- org/10.1071/SH14220. Australia: Annual report of trends in behaviour lance among MSM in 8. UNAIDS/WHO. Second Auckland, New Zealand generation HIV Surveil- 2014 (Monograph 7/2014). 2002-2011: condom use lance for HIV: the next Sydney: Centre for Social stable and more HIV decade. Geneva: UNAIDS/ Research in Health, testing. Sex Transm WHO Working Group UNSW Australia, 2014. Infect. 2014;90:133-8. on Global HIV/AIDS and 12. Finlayson T, Binh L,

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 60 www.nzma.org.nz/journal ARTICLE

Smith A, et al. HIV Risk, New Zealand HIV Clinical 25. Ludlam A, Saxton P, Prevention, and Testing Update, 8th May 2015. Dickson N, Hughes A. Behaviors Among Men http://player.vimeo.com/ Attitudes towards safe sex Who Have Sex With Men: external/130713756. among men who have sex National HIV Behavioral sd.mp4?s=eea703dd- with men in New Zealand: Surveillance System, b3ed5aa75178674d7d- Findings from the GAPSS 21 U.S. Cities, United ace69a&profile_id=112 and GOSS surveys 2002- States, 2008. MMWR. 19. Cohen MS, Chen YQ, 2011. Research brief to 2011; 60(SS14);1-34. McCauley M, et al. Preven- the Ministry of Health. 13. Williamson L, Dodds J, tion of HIV-1 infection Dunedin: AIDS Epidemiol- Mercey D, Hart G, et al. with early antiretroviral ogy Group, University of Sexual risk behaviour therapy. N Engl J Med. Otago, 2012. https://www. and knowledge of HIV 2011; 365:493-505. fmhs.auckland.ac.nz/ status among commu- 20. The INSIGHT START assets/fmhs/soph/sch/ nity samples of gay Study Group. Initiation gmsh/docs/gapss-goss-atti- men in the UK. AIDS. of antiretroviral therapy tudes-research-brief.pdf 2008; 22:1063-1070. in early asymptomatic 26. Lachowsky NJ, Dewey CE, 14. Truong H, Kellogg T, HIV infection. N Engl J Dickson NP, et al. Habitual Klausner J, et al. Increases Med. 2015; 373:795-807. condom use across partner in sexually transmitted 21. McCormack S, Dunn D. type and sexual position infections and sexual Pragmatic Open-Label among younger gay and risk behaviour without Randomised Trial of bisexual men: findings a concurrent increase Preexposure Prophylaxis: from New Zealand HIV in HIV incidence among The PROUD Study. 2015 behavioural surveillance men who have sex with Conference on Retrovi- 2006-2011. Sex Transm men in San Francisco: ruses and Opportunistic Infect. 2015; 91:445-450 a suggestion of HIV Infections (CROI), Seattle, 27. Coughlan E, Young H, serosorting?. Sex Transm USA, abstract 22LB, 2015. Parkes C, et al. A novel Infect. 2006; 82:461-466. 22. Molina JM, Capitant C, response to an outbreak 15. Crawford J, Rodden P, Charreau I, et al. On of infectious syphilis Kippax S, Van de Ven Demand PrEP With Oral in Christchurch, New P. Negotiated safety TDF-FTC in MSM: Results Zealand. Sex Health. and other agreements of the ANRS Ipergay Trial. 2015; 12:170-3. between men in rela- 2015 Conference on Retro- 28. The Institute of Envi- tionships: risk practice viruses and Opportunistic ronmental Science and redefined. Int J STD Infections (CROI), Seattle, Research Ltd. Enhanced AIDS. 2001; 12: 164-170. USA, abstract 23LB, 2015. Surveillance of Infectious 16. Grulich A, de Visser 23. Kippax S. Biomedical Syphilis in New Zealand R, Badcock P, et al. prevention: rhetoric and Sexual Health Clinics – Homosexual experience reality. HIV Australia. 2013. Porirua: ESR, 2014. and recent homosexual 2015; 13:8-11. https:// 29. Basu I, Bromhead C, Balm encounters: the Second www.afao.org.au/library/ Australian Study of Health hiv-australia/volume-13/ M, et al. Lymphogranu- and Relationships. Sex vol-13-number-2-horizons/ loma venereum in men Health. 2014; 11:439-450. biomedical-preven- who have sex with men: 17. Lucassen M, Clark tion-rhetoric-and-reality evidence of local transmission in New Zealand. T, Moselen E, et al. 24. National AIDS Trust. HIV N Z Med J. 2015;128:25-9. Youth’12 The Health and prevention in England’s Wellbeing of Secondary high prevalence local 30. Public Health England. School Students in New authorities 2013/14 and Sexually transmitted Zealand: Results for Young 2014/15: Examining infections and chlamydia People Attracted to the spending on primary HIV screening in England, Same Sex or Both Sexes. prevention and additional 2014. Health Protection Auckland: The University HIV testing. London: Report. 2015; 9:22. of Auckland, 2014. National AIDS Trust, https://www.gov.uk/ 18. Rich J. HIV prevention 2015. http://www.nat. government/uploads/ among gay and bisexual org.uk/media/Files/ system/uploads/attach- men. Paper presented at Publications/Preven- ment_data/file/437433/ Three Decades of HIV in tion_Report_2015-1.pdf hpr2215_STI_NCSP_v6.pdf

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 61 www.nzma.org.nz/journal ARTICLE

Drug misuse in sport: a New Zealand perspective Andrew Curtis, David Gerrard, Peter Burt, Hamish Osborne

ABSTRACT AIMS: Drug misuse in elite sport is a world-wide phenomenon. This article explores the culture of contemporary sport, provides estimates of doping prevalence, discusses dietary supplementation and highlights major factors influencing high-performance athletes and their support personnel. The aim is to stimulate discussion, informed by the World Anti-Doping Code (WADC), which is particularly relevant to doctors caring for athletes. METHODS: Online databases were searched for relevant peer-reviewed research from 2009 to 2015. Comparative New Zealand data have been included. RESULTS: Estimates of the prevalence of sports doping range from less than 1% to as high as 52%, dependent upon the demographics of the identified cohort. The culture of elite sport, personal stressors, competitive demands, financial reward and the influence of an ‘entourage’ of support personnel were identified as critical determinants of drug misuse. CONCLUSIONS: The culture of elite contemporary sport is seductive to many aspiring young athletes. To combat drug misuse, effective education should embody moral, ethical and clinical dangers, recognising the importance of support at times of increased athlete vulnerability. Inadvertent doping from product contamination is a recognised risk of unsupervised dietary supplementation. Doctors responsible for the care of high-performance athletes must be cognisant of these issues and the provisions of the WADC.

he use of drugs to enhance sports This article references the body of research performance is a global phenomenon on doping prevalence in sport, discussing Tthat continues to receive wide media its associated culture, common reasons, attention. The number, variety and use key personnel and prevention strategies of legal and illegal drugs has increased in to assist medical professionals in the New 1 recent years. Drug misuse in elite sport is Zealand context. monitored internationally by the World An- ti-Doping Agency (WADA), while Drug-Free Sport New Zealand (DFSNZ) is responsible Review methodology for national athlete testing and education. MEDLINE, PubMed, Scopus and SPORT- Discus online databases were searched Recently, it has been proposed that for peer-reviewed research from 2009 young athletes transition incrementally to January, 2015, using combinations from their use of ‘permitted’ to illegal substances, with the suggestion of ‘harm of the terms, “doping”, “performance minimisation’ as an approach to counter enhancing drug*”, “performance enhancing this.1 Health professionals, particularly substance*”, “drugs”, “anti-doping”, doctors, are traditionally recognised by “education”, “sport*”, “prevalence”, athletes as a trusted resource for all matters “prevention”, “athlete*”, “elite”, “sub-elite,” of drug efficacy and safety, including the “recreational” and “adolescent.” A second- use of dietary supplements and perfor- stage ‘snowball’ search scanned reference mance-enhancing agents.2 However, many lists of published articles for relevant physicians lack fundamental knowledge manuscripts and considered articles recom- to provide adequate advice to athletes.3 mended by the online databases at the time

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 62 www.nzma.org.nz/journal ARTICLE

the original manuscripts were downloaded. questions and the obvious sensitivity of A total of 232 references were identified the responses. International drug testing that related to the prevalence, prevention results, collated by WADA, demonstrate an and culture of drug misuse in sport. Only approximate prevalence of 2% positive tests six of these were specific to New Zealand. per year. However, the true prevalence is thought to be closer to 10%5 with a recent Results and review of the literature yielding an estimation of 14–39%.4 discussion In 2013/14, DFSNZ carried out 925 drug Definitions tests on elite athletes, with 4 violations, For the purposes of this review, ‘doping” a doping prevalence of 0.004%, which is defined as a breach of the WADA rules, compares to 0.007% in 2012/13 and 0.005% including use or attempted use of a in 2011/12.6 However, these tests do not prohibited substance or method. These include gym users or athletes not subjected include anabolic agents, peptide hormones, to anti-doping regulations. Another paper stimulants, diuretics, beta-2 agonists reported 5 of 32 New Zealand body builders and recognised performance-enhancing admitting the use of androgenic anabolic methods, such as blood and gene doping, steroids (AAS) at some stage in their career.6 urine tampering or intravenous infusions Australian-based studies of doping amongst unless medically indicated. The use of illicit elite athletes reported an 8% response, “recreational” drugs, including narcotics compared with 52% in male gym users,7,9 and cannabinoids, is also considered a while 25% of a Canadian cohort of junior breach of the WADC. Contemporary liter- provincial athletes reported PED use in ature uses the interchangeable terms the previous year,9 and up to 12% of an “performance-enhancing drugs” (PEDs), American high school student cohort “banned drugs” and “doping” with resulting reported AAS use.10,11 From these estimates, confusion.1,4 In this review, doping infers the 10% of athletes seen by a health profes- collective of PEDs, performance-enhancing sional are possibly using a PED, and 1 in methods and illicit drugs. 3 are at risk of inadvertent doping from Breaches of the WADA rules are supplement use. considered either intentional or inad- Drug-User Profile vertent. The former implies ‘cheating’, Competitive athletes who intentionally whereas the latter may result from dope are categorised as “…villains, 1 supplement contamination or ignorance. mavericks and professionals.”1 “Villains” Acts of doping in sport focus primarily cheat deliberately, while “mavericks” on an intent to enhance performance, display an ignorant disregard for the rules. while ‘inadvertent doping’, not generally “Professionals” however—purported to considered as purposeful, is deemed a be the largest group—progress from diet consequence of either unknown product and lifestyle changes, to supplementation, contamination or ‘recreational’ drug use. and finally to banned substance use. It is However, the WADC applies strict personal argued that these athletes are not ‘cheaters’, liability to drug misuse, making no such but products of the intensely compet- discrimination when considering violations. itive, commercialised world of elite sport, As a result, any athlete found ‘positive’ may whereby they are driven to train with be sanctioned in accordance with options greater intensity for longer periods. from a reprimand, to the rarely used At a recreational level, so-called lifetime ban from sport. For the purposes of ‘gym users’ plus ‘power and strength this review, the authors use ‘doping’ to refer sportspeople’ are more likely to use AAS to deliberate, banned drug-use and ‘inad- or growth hormone derivatives.12,13 This vertent doping’ for product contamination systematic review of anabolic steroid or ignorance of the rules. use listed appearance, aggression or Prevalence enhanced performance as the most A true prevalence of doping in sport is relevant reasons for doping.13 These dopers difficult to determine given the limitations were characterised as being male, under of data collection, the intrusiveness of the 30 years of age, mistrusting of medical

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 63 www.nzma.org.nz/journal ARTICLE

professionals and with comorbidities also emphasised, and for 292 New Zealand including depression and a history of athletes interviewed, coaching style was a illicit drug use.12,13 Furthermore, female determinant in an increased athlete doping AAS-users have a much higher risk of risk.19 This influence was also reflected by dependency than male counterparts.14 studies of elite Scottish,27 German,16 and Therefore, recreational sportspeople with Greek28 athletes. the characteristics described should prompt Culture of sport medical professionals to be wary of their While the culture of sport has been iden- potential for drug misuse. tified as shaping an athlete’s attitudes and Times when athletes are at an intentions to dope, the public and the media increased doping risk consider doping as simply another form The culture of doping is as varied as of ‘cheating’.13,29 Athletes caught cheating the sports, sub-cultures, ability, ages and are commonly portrayed as ‘bad’, with personalities of the users.1,2,15-19 Notwith- the role of their entourage often ignored,28 standing, some individual characteristics despite compelling evidence that they and specific determinants have emerged are complicit.17 Athletes are frequently that could assist doctors who regularly ‘villainised’ when caught using drugs in a manage athletes. As a group, athletes have recreational setting. Multiple Olympic gold been identified as being more likely to medallist Michael Phelps was publicly chas- use a PED if offered the chance.20,21 Qual- tised for his one-time use of cannabis,30,33 yet itative research involving a cohort of Barack Obama as a Presidential candidate 147 UK athletes identified reliability, rule was praised for honesty in declaring his abiding and role modelling as ‘protective youthful, cannabis and cocaine use.31 behaviours’, while rule breaking, bad Elite athletes are more likely to dope temperament and a win-at-all-costs attitude if they believe that other athletes are were risk factors for doping.22 An athlete’s doping.21,32,33 For example, eight elite and ‘doping risk’ was also reported to increase neo-elite cyclists, interviewed prior to during critical events, such as selection/ turning professional, viewed doping as de-selection,23,24 during recovery from cheating, yet once they became professional injury and when negotiating crucial spon- they regarded doping as an inevitable sorship deals.17 These transitions were progression in performance enhancement. considered to be times of psychosocial They also claimed elite sport as being challenge with an enhanced risk of doping. deleterious to health, rationalising that At such times, social support, individual PEDs conferred a protective influence.32,34 coping mechanisms2,25 and the influence of Boundaries can be blurred between medical advice was deemed critical.3 ‘legitimate’ performance enhancement, Entourage—influence and including physiological testing, nutri- knowledge tional supplementation or biomechanical computer-modelling and frank doping to A complex of individuals, identified compensate for media pressures, spon- as the ‘athlete entourage’, contributes to sorship or public expectation.35,36 Times of the environment of every elite athlete. increased vulnerability demand concerted Doctors, coaches, trainers, family, friends, education and awareness from all stake- teammates and physiotherapists are holders, particularly doctors. acknowledged sources of knowledge, leadership and support.2 Yet a study of the Dietary supplements anti-doping knowledge of 292 Australian Dietary supplementation in sport is support personnel revealed that 40% had no common, with the internet, team mates, specific training, despite providing advice coaches and athletic trainers providing the to athletes.26 This study also revealed that most common sources of information.37 An 32% of these support personnel ignored the unpublished survey of elite New Zealand unethical behaviour of colleagues, despite a athletes reported a 93% usage of 3 supple- WADC obligation to report doping offences ments in the prior 6 months,38 findings irrespective of confidentiality.3 The impor- comparable to data from a similar Canadian tance of the coach in the social network is study.37 Inadvertent doping is a potential

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 64 www.nzma.org.nz/journal ARTICLE

consequence of supplementation, with of “…displacement or diffusion of responsi- products frequently not subjected to strict bility, advantageous comparison, distortion manufacturing and quality control. Fifteen of consequences, moral justification and percent of internet-sourced supplements euphemistic labelling”.13 These phenomena have been reported with steroid contam- are documented in body builders,7,13,35 ination39 as well as potent psychoactive weightlifters,23 cyclists,34 and in 1,188 substances, including DMBA (1,3-dimeth- Australian adolescents were predictive ylbutylamine) and its analogues.39 Dietary of doping attitudes, regardless of social supplement users are also shown to be demographics or athletic status.33 In order at greater risk of doping than non-users, to counter forms of moral disengagement, reflected in studies of elite UK athletes,15 medical professionals must recognise the Australian and Greek high school process and develop appropriate counter 19,25 24,36 students, amateur Australian cyclists, arguments.35 and Croatian rugby players.41 A more permissive attitude towards doping has mirrored increasing supplement use, with Conclusions recovery from injury or training, improved Despite the importance of sport in our performance, increased muscle size and society, there is a dearth of New Zealand body image as common reasons.15,35,37 research relating to sports doping. Interna- Sources of supplements and reasons for tional figures suggest that doping is more their use are matters for doctors to explore common than figures would suggest and with athletes in their care. that deterrence through punitive measures alone is ineffective. Body image and moral An understanding of drug misuse in disengagement sport deserves a wider, empathetic view ‘To look good,’ is an oft-cited reason for that embodies the culture of sport and recreational athletes, especially serious gym the influence of the athlete entourage of users, to use AAS and supplements.42 Both support personnel. The most common AAS and supplement use are reportedly reasons given for PED use are to improve associated with an increased alcohol and looks, increase performance, to cope with illicit drug consumption,13,42 low self-esteem the demands of training, or to recover or a negative body image, and participation in sports where muscle bulk is important.10 from injury. More recent research also Product source is important, with 50–75% of suggests impressionable young athletes PEDs being reportedly purchased online.21 may see doping as a natural progression of One study used laboratory testing of 57 AAS performance enhancement and be willing or growth hormone derivatives purchased to risk sanctions and personal health in online and reported 42% being either the pursuit of success. Regardless, athletes contaminated with bacteria, containing no taking supplements or PEDs bought online active anabolic ingredient or raising other risk their health through possible contam- safety issues.21 The same study reported ination. Effective educational strategies that testing 634 nutritional supplements encourage themes of health, morality and found many to contain some trace of AAS. refusal skills, while acknowledging that The potential co-morbidities and risks for there are periods of increased athlete PEDs or supplements purchased online vulnerability. Medical professionals in is important information for all medical particular need to be increasingly wary of professionals, but particularly doctors, to be these times of increased risk. aware of. Doctors treating competitive or recreational Athletes frequently rationalise doping on athletes carry a burden of responsibility spurious grounds that ignore health and in their knowledge of dietary supplemen- safety.23,32 A strategy known as ‘moral disen- tation and prohibited substances that gagement’ negates the immoral actions of reflects patient health and the spirit of sport cheating through established mechanisms embodied in the World Anti-Doping Code.

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 65 www.nzma.org.nz/journal ARTICLE

Competing interests: Dr. Gerrard reports he is currently the Chair of the World Anti-Doping Agency (WADA) Therapeutic Use Exemption Committeeand a member of the WADA Health Medicine and Research Committee, both voluntary positions. Author information: Andrew N Curtis, University of Otago, Christchurch; David Gerrard, Medicine, Dunedin School of Medicine, University of Otago, Dunedin; Peter Burt, Dunedin School of Medicine, University of Otago, Dunedin; Hamish Osborne, Dunedin School of Medicine, University of Otago, Dunedin Corresponding author: David Gerrard, Medicine, Dunedin School of Medicine, University of Otago, Dunedin [email protected] URL: www.nzma.org.nz/journal/read-the-journal/all-issues/2010-2019/2015/vol-128-no-1426-4- december-2015/6747

REFERENCES: 1. Petróczi A. The doping 7. Probert A, Leberman S. ic-androgenic steroid use mindset—Part I: Impli- The Value of the Dark Side: initiation: A systematic cations of the Functional An Insight into the Risks review and synthesis Use Theory on mental and Benefits of Engaging of qualitative research. representations of in Health-compromis- Substance Abuse: doping. Performance ing Practices from the Treatment, Prevention, Enhancement & Health. Perspective of Competitive and Policy. 2014;9(1). 2013;2(4):153-63. Bodybuilders. European 13. Boardley ID, Grix J, Dewar Sport Management Quar- 2. Chan D, Hardcastle S, AJ. Moral disengagement terly. 2009;9(4):353-73. Lentillon-Kaestner V, and associated processes Donovan R, Dimmock J, 8. Dunn M, Thomas JO, in performance-enhanc- Hagger M. Athletes’ Beliefs Swift W, Burns L. Elite ing drug use: a national About and Attitudes athletes’ estimates of qualitative investigation. Towards Taking Banned the prevalence of illicit Journal of Sports Sciences. Performance-Enhancing drug use: Evidence for 2014;32(9):836-44. the false consensus Substances: A Quali- 14. Ip EJ, Barnett MJ, Tene- effect. Drug and Alcohol tative Study. 2014. rowicz MJ, Kim JA, Wei Review. 2012;31(1):27-32. H, Perry PJ. Women 3. Backhouse S, McKenna J. 9. Goulet C, Valois P, Buist A, and anabolic steroids: Doping in sport: A review Cote M. Predictors of the An analysis of a Dozen of medical practitioners’ Use of Performance-En- users. Clinical Journal knowledge, attitudes and hancing Substances by of Sport Medicine. beliefs. Int J Drug Policy. Young Athletes. Clinical 2010;20(6):475-81. 2011;22(3):198-202. Journal of Sport Medi- 15. Backhouse S, Whitaker 4. de Hon O, Kuipers H, cine. 2010;20(4):243-8. L, Petroczi A. Gateway van Bottenburg M. 10. Dandoy C, Gereige RS. to doping? Supplement Prevalence of doping Performance-enhancing use in the context of use in elite sports: a drugs. Pediatrics in preferred competitive review of numbers Review. 2012;33(6):265-72. situations, doping attitude, and methods. Sports 11. Thorlindsson T, Hall- beliefs, and norms. Medicine. 2014:1-13. dorsson V. Sport, and Scand J Med Sci Sports. 5. Batt A. The role of UK use of anabolic andro- 2013;23(2):244-52. Anti-Doping in the fight genic steroids among 16. Sas-Nowosielski K, against doping in sport. Icelandic high school Swiatkowska L. Goal International Journal of students: a critical test orientations and attitudes Sport Policy and Politics. of three perspectives. toward doping. Int J Sports 2011;3(2):261-70. Subst Abuse Treat Prev Med. 2008;29(7):607-12. 6. DFSNZ. Annual Report Policy. 2010;5:32. 17. Smith ACT, Stewart 2013/2014. Auckland, 12. Sagoe D, Andreassen CS, B, Oliver-Bennetts S, New Zealand: Drug Free Pallesen S. The aetiology McDonald S, Ingerson L, Sport New Zealand, 2014. and trajectory of anabol- Anderson A, et al. Contex-

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 66 www.nzma.org.nz/journal ARTICLE

tual influences and athlete club cycling: a life-course 32. Lentillon-Kaestner V, attitudes to drugs in analysis. Scandinavian Hagger MS, Hardcastle sport. Sport management Journal of Medicine S. Health and doping review. 2010;13(3):181-97. & Science in Sports. in elite-level cycling. 18. Yager Z, O’Dea JA. 2013;23(6):e361-e72. Scandinavian Journal of Relationships between 25. Barkoukis V, Lazuras L, Medicine & Science in body image, nutritional Lucidi F, Tsorbatzoudis H. Sports. 2012;22(5):596-606. supplement use, and Nutritional supplement 33. Skinner J, Moston S, attitudes towards doping and doping use in sport: Engelberg T. The relation- in sport among adolescent Possible underlying social ship between moral code, boys: implications for cognitive processes. participation in sport, prevention programs. Scandinavian journal and attitudes towards J Int Soc Sports Nutr. of medicine & science performance enhancing 2014;11(13):1-8. in sports. 2015. drugs in young people. Montreal: World Anti-Dop- 19. Hodge K, Hargreaves 26. Mazanov J, Backhouse ing Agency (WADA), 2012. EA, Gerrard D, Lons- S, Connor J, Hemphill D, dale C. Psychological Quirk F. Athlete support 34. Lentillon-Kaestner V. The Mechanisms Underlying personnel and anti-dop- development of doping Doping Attitudes in Sport: ing: Knowledge, attitudes, use in high-level cycling: Motivation and Moral and ethical stance. From team-organized Disengagement. Journal of Scandinavian Journal of doping to advances in Sport & Exercise Psychol- the fight against doping. Medicine & Science in ogy. 2013;35(4):419-32. Scandinavian Journal of Sports. 2014;24(5):846-56. 20. Dodge T, Stock M, Litt D. Medicine & Science in 27. Allen J, Taylor J, Dimeo Judgments About Illegal Sports. 2011;23(2):189-97. P, Dixon S, Robinson L. Performance Enhancing 35. Engelberg T, Moston Predicting elite Scottish Substances: Reasoned, S, Skinner J. The final athletes’ attitudes Reactive or Both? Journal frontier of anti-doping: towards doping: exam- of Health Psychology. A study of athletes who ining the contribution of 2012;18(7):962-71. have committed doping achievement goals and violations. Sport Manage- 21. Graham MR, Ryan P, motivational climate. ment Review. 2014. Baker JS, Davies B, Journal of Sports Thomas N-E, Cooper Sciences. 2014:1-8. 36. Outram SM, Stewart B. S-M, et al. Counterfeiting Condemning and condon- 28. Barkoukis V, Lazuras in performance- and ing: Elite amateur cyclists’ L, Tsorbatzoudis H, image-enhancing drugs. perspectives on drug use Rodafinos A. Motivational Drug Testing and Anal- and professional cycling. and sportspersonship ysis. 2009;1(3):135-42. Int J Drug Policy. 2015. profiles of elite athletes 22. Whitaker L, Long J, 37. Lun V, Erdman KA, Fung in relation to doping Petróczi A, Backhouse TS, Reimer RA. Dietary behavior. Psychology SH. Athletes’ percep- supplementation practices of Sport and Exercise. tions of performance in Canadian high-per- 2011;12(3):205-12. enhancing substance formance athletes. user and non-user 29. Mazanov J, Huybers T, Int J Sport Nutr Exerc prototypes. Performance Connor J. Prioritising Metab. 2012;22(1):31-7. health in anti-doping: Enhancement & Health. 38. Hellemans I, MacDonald S, What Australians think. 2012;1(1):28-34. Skidmore P. Use of Dietary Journal of Science and 23. Kirby K, Moran A, Guerin Supplements in Elite Medicine in Sport. S. A qualitative analysis New Zealand Athletes. 2012;15(5):381-5. of the experiences of Dunedin: Department elite athletes who have 30. Macur J. Phelps Disci- of Human Nutrition, admitted to doping for plined Over Marijuana University of Otago, performance enhance- Pipe Incident. New York 2009 2009. Report No. ment. International Times. 2009 Feb 6, 2009. 39. Geyer H, Parr MK, Koehler Journal of Sport Policy and 31. Seelye KQ. Barack K, Mareck U, Schanzer Politics. 2011;3(2):205-24. Obama, asked about W, Thevis M. Nutritional 24. Stewart B, Outram S, Smith drug history, admits he supplements cross-con- ACT. Doing supplements to inhaled. New York Times. taminated and faked improve performance in 2006 Oct 24, 2006. with doping substances.

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 67 www.nzma.org.nz/journal ARTICLE

J Mass Spectrom. Drug Testing and Anal- 2014;22(3):226-39. 2008;43(7):892-902. ysis. 2015;7(1):83-7. 42. Dodge T, Hoagland MF. 40. Cohen PA, Travis JC, 41. Sekulic D, Bjelanovic The use of anabolic Venhuis BJ. A synthetic L, Pehar M, Pelivan K, androgenic steroids and stimulant never tested Zenic N. Substance use polypharmacy: A review in humans, 1,3-dimeth- and misuse and potential ylbutylamine (DMBA), doping behaviour in rugby of the literature. Drug is identified in multiple union players. Research and Alcohol Dependence. dietary supplements. in Sports Medicine. 2011;114(2-3):100-9.

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 68 www.nzma.org.nz/journal ARTICLE

‘Real-time’ burden of community and healthcare- related infections in medical and rehabilitation patients in a public hospital in Auckland, New Zealand Kerry Read, Hasan Bhally

ABSTRACT AIMS: To determine the prevalence and spectrum of infections on admission, or acquired during hospitalisation (HAI) at Waitakere Hospital, Auckland. METHODS: A questionnaire was completed on two separate days for all adult in-patients admitted to medical and rehabilitation wards for greater than 24 hours. Information obtained included patient characteristics, the presence and type of infection on admission or acquired during hospitalisation, as well as information on indwelling devices. RESULTS: Infection was the admitting diagnosis in 81 (41%) of 195 patients reviewed, with lower respiratory tract infection (LRTI) diagnosed in 50%, urine infections in 22% and cellulitis 18%. Only 40% LRTIs were supported by radiology or microbiological criteria. Twenty-five HAIs occurred in 21 patients (cumulative and point prevalence of 10.7% and 5.0% respectively). Urinary tract infection (UTI) was the most common HAI in 13 patients (62%), including 4 catheter-related infections. Patients with HAI were older and appeared to have had longer hospital stays, and higher urinary catheter usage. CONCLUSIONS: This study highlights the ongoing high burden of infections contributing to hospitalisation of adult patients in a developed country. The prevalence of HAI, patient characteristics and risk factors are comparable to previous studies in similar settings.

nfection, either on admission to hos- The main purpose of our study was to pital or acquired during inpatient stay determine the prevalence and spectrum of I(hospital-acquired infection, HAI), con- infections at the time of hospital admission tinues to be a common diagnosis requiring or those acquired during hospitalisation. healthcare provision. Baker et al recently The study population were adult medical reported that in New Zealand, infection-re- patients admitted to general medical or 1 lated admissions to hospital were rising. In rehabilitation wards at Waitakere Hospital. New Zealand, it is estimated that about 10% This 135-bed hospital provides acute of inpatients will develop an HAI sometime general medical and rehabilitation services during their hospital admission.2 HAIs to a population of approximately 250,000 result in additional costs largely related to in West Auckland. There are 92 general longer lengths of stay, as well as extra costs medical and 45 rehabilitation beds. for diagnostic tests and treatment. Graves et al, in 2003, estimated that the annual cost to New Zealand hospitals from HAIs in medi- Methods cal and surgical patients is between $50 and A survey of all adult patients was $85 million.3 performed on two separate days, in late

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 69 www.nzma.org.nz/journal ARTICLE

winter (29 August) and late spring (19 symptoms, treated by clinician as UTI, or (c) November), 2013. All 6 medical and 3 Mixed culture of >105 CFU/ml and pyuria in rehabilitation teams were instructed to a symptomatic patient treated as UTI); skin complete a study questionnaire incorpo- and soft tissue (SSTI) (cellulitis, abscess, rating questions on patient demographics, infected ulcers); Influenza-like illness (ILI) co-morbidities, the reason for admission (acute onset fever with headache, myal- and presence of infection on admission, gia’s, sore throat, coryza); gastrointestinal information about indwelling devices, and (including C. difficile, gastroenteritis); and details about any HAI present either at the others. time of evaluation, or occurring at any stage This study was registered and approved during the current hospitalisation. by the Awhina Knowledge and Research HAI was diagnosed if an infection Centre of the Waitemata District Health occurred on or after the third calendar day Board. of hospitalisation (where admission day Statistical analysis was performed using is considered first calendar day), as per Microsoft Analyse-it software. standard CDC/NHSN definition.http://www. cdc.gov/nhsn/PDFs/pscManual/2PSC_Identi- Results fyingHAIs_NHSNcurrent.pdf A total of 271 patients were in hospital Patients admitted for less than 24 hours during the two study days, including 49 new were excluded from the study. Admission admissions. Hospital occupancy was 99% to a rehabilitation ward was considered a on both days. One hundred and ninety-five continuum of the current hospitalisation patients (98 in August, 97 in November), if patients were transferred from another consisting of 127 medical and 68 reha- inpatient service. bilitation-ward patients were included. The diagnosis of infection were based on Seventy-six patients had been in hospital clinical criteria by the respective medical less than 24 hours and were excluded. Data teams and microbiologic and/or radiologic from both periods were combined. The confirmation was not required (except in mean patient age was 73.5 years (range bloodstream infections, urinary tract infec- 17–98 years), and females predominated tions and C. difficile). These responses were (65%). Common co-morbidities included used to calculate outcome measures-point ischaemic heart disease (n=63), diabetes prevalence and cumulative prevalence. (n=50) and chronic lung disease (n=32). Point prevalence was defined as presence Fifteen patients had advanced dementia. of an active HAI at the time of performing Ethnicity data revealed that 72% of respon- the survey. dents were New Zealand Europeans, 12% Cumulative prevalence was defined Pacific peoples, 3.6% Māori and 3.6% Asian. as HAI diagnosed at any time during the Data were missing for 8.7% of participants. current hospital stay (including patients in Patients were predominantly admitted from point prevalence analysis). the community (75%) or long-term care facilities (20%). Denominator for both measures was the total number of inpatients present on the The length of stay before the study day respective survey days. assessment was available only for the August cohort. This was a mean of 5.5 days Infection was categorised as lower (range 2–29 days) for medical patients and respiratory tract (LRTI) (including acute 12.5 days (range 2–30 days) for rehabili- bronchitis, COPD exacerbation, clinical- tation patients. ly-suspected pneumonia with or without No significant differences in demographic the presence of new infiltrate on chest characteristics were observed between the x-ray, exacerbation of bronchiectasis, etc); August and November cohorts. urinary tract (UTI) (uncomplicated cystitis to urosepsis with (a) pyuria, with or without Infection as admitting diagnosis fever or urinary symptoms, and growth of Overall, 85 episodes of infection were at least 103 CFU/ml of up to two different documented in 81 (41%) patients at the bacteria, or (b) pure growth of single time of admission; 71 as the primary bacterial species >105 CFU/ml, regardless of reason for admission, and 10 as secondary

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 70 www.nzma.org.nz/journal ARTICLE

Figure 1: Spectrum of 85 infections seen on hospital admission

LRTI: lower respiratory tract infection, UTI- urinary tract infection

Table 1: Characteristics and spectrum of hospital acquired infections in 21 patients

No. Age Reason for admission HAI type Location at Duration of Indwelling (Yr.) acquisition stay before device HAI (days) 1 91 Rehab. post hip arthroplasty UTI Rehab. 2 None

2 93 Reduced mobility UTI Rehab 10 None

3 81 Stroke, UTI LRTI-HA Gen Med 2 None

4 80 Pneumonia UTI Gen Med 10 IDC

5 94 Mechanical fall UTI Rehab 17 IDC, IVL

6 83 UTI, pneumonia, cellulitis C. difficile Gen Med 40 IDC, IVL

7 81 UTI LRTI-HA Gen Med 4 IVL

8 98 Cellulitis LRTI- HA Rehab 5 IDC

9 70 Reduced mobility LRTI-HA Gen Med 4 IDC

10 88 Cardiac Wound Rehab 48 IDC

11 85 Collapse Wound PPM UTI Gen Med 20 IVL

12 88 COPD exacerbation UTI Gen Med 11 IDC

13 84 Cellulitis UTI Gen Med 25 IDC

14 81 Rehab post hip arthroplasty LRTI- HA Rehab 6 IVL

15 87 Pneumonia Wound—pressure sore Rehab 31 IVL

16 79 Rehab post hip fracture UTI Rehab 1 IVL

17 N/A N/A UTI N/A N/A N/A

18 85 Collapse UTI Rehab 8 None

19 90 Fall UTI Gen med 6 IVL

20 88 Mobility issues UTI Rehab 14 None

21 86 Fall UTI, LRTI-HA Rehab 21 None

LRTI-HA: Hospital acquired lower respiratory tract infection, UTI: urinary tract infection, PPM: permanent pacemaker, IVL: peripheral intravenous line, IDC: indwelling urinary catheter, N/A: not available. Gen Med: General medicine, Rehab: Rehabilitation ward.

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 71 www.nzma.org.nz/journal ARTICLE

Table 2: Selected baseline clinical characteristics and overall duration of hospital stay in patients with and without hospital-acquired infections.

HAI Non-HAI n=21 n=174 Average age (years) 85.6 75.2

Male sex 28% 40%

Average length of total stay (days) 32.9 15.1

Location in rehabilitation ward at time of HAI 71% 29%

% with indwelling devices prior to HAI 43% 46% diagnosis (n=89) (n=9) (n=80)

% with invasive procedures prior to HAI 19% 6.8% diagnosis (n=16) (n=4) (n=12)

diagnosis. Lower respiratory tract and present on admission. Twelve of 66 infections were the most prevalent form patients (18%) in the rehabilitation ward of infection (Figure 1). Other common had indwelling devices compared to 76 of infections on admission included urinary 127 medical patients (60%). tract infections and cellulitis. Influenza-like Table 2 shows selected characteristics of illness were only seen in August. patients diagnosed with an HAI compared Notably, of the 35 patients with primary to the non-HAI group. We offer this table admitting diagnosis of LRTI, only 14 had a for descriptive purposes only, as we cannot definite indication for antibiotic use ie, chest draw conclusions on any differences x-ray with consolidation (11) or positive between these groups without control of sputum culture/urinary antigen without confounding or mediating factors. Patients consolidation (3). with HAI were older and more likely present Hospital-acquired infections in rehabilitation ward at the time of diag- Twenty-one patients developed 25 nosis. Average length of stay was much episodes of HAI, identified between longer in HAI group, but the study was admission and study dates (cumulative not designed to assess whether this was a HAI prevalence of 10.7%). Ten patients had risk factor for the HAI or a consequence an HAI present on the study days (point of the HAI. A high proportion of patients prevalence 5%). Information on demo- (43% and 46%) had either an indwelling graphics and device use was unavailable for peripheral vascular catheter or urinary one patient. Table 1 shows the individual catheter present prior to diagnosis of HAI. In characteristics of patients and spectrum of contrast, only 16 (8%) had invasive proce- HAIs. The mean age of patients was 85 years dures performed (19% in HAI and 6.8% in and total length of stay 33 days. The mean non-HAI group). There were no immediate duration of stay prior to diagnosis of HAI post-surgical patients in the study cohort. was 14.5 days (range 1–48 days). None of the 4 HAI’s were directly related to Urinary tract infections were the most the procedure performed. No difference was common HAI in 13 patients (62%), and found in terms of presence of indwelling included two cases of urosepsis. Other devices or invasive procedures in the HAI vs HAIs included six hospital acquired LRTIs, non-HAI group. three wound infections and one case of C. difficile colitis. Approximately 60% of HAI Discussion events with positive cultures yielded E. coli Infections are a common reason for or Klebisella sp. (including ESBL-producing admission to hospital. In addition, hospital Enterobacteriaceae). acquired infections can complicate inpatient An indwelling device was present in 89 stay. Our study shows that in adult patients patients (45.6%); 75 had peripheral intra- hospitalised for more than 24 hours, clinically venous lines and 12 had urinary catheters. diagnosed infections were present in 41% at Eight of 12 urinary catheters were long-term the time of admission either as a primary or

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 72 www.nzma.org.nz/journal ARTICLE

secondary diagnosis. Lower respiratory tract of medical and surgical patients (excluding infections were the most common type of those on renal replacement therapy) was in infection present on admission. excess of $20,000 per case. The main factor Baker et al1 showed that infections, as a contributing to the additional cost was the principle diagnosis based on ICD diagnostic longer length of stay. codes, accounted for 27% of hospitalisations Our study was performed by utilising in New Zealand between 2004 and 2008. the services of training doctors who, after They included admissions with durations of an initial information session with the less than 24 hours and paediatric patients, study investigators about data collection, and also noted increasing rates of admis- completed the survey questions for their sions related to infection from 1989 to 2008. respective team patients in a real time. This The published New Zealand data docu- approach minimised recall or interpretation menting the rates of infection present at bias. However, several limitations need hospital admission in adults is scarce. mention. Firstly, we assessed the burden of Hospital-acquired infections are common. infection in adult patients only admitted to a It is estimated that about 10% of patients medium-sized Auckland hospital where no admitted to a hospital in New Zealand, or any acute surgical services are present. Results developed country, will acquire a health- may therefore not be generalised to other care-related infection.3 In our study, HAIs centres. Secondly, our exclusion of patients occurred with a cumulative prevalence of with less than 24-hour stay may have 10.7% and point prevalence of 5%. The HAI underestimated infections diagnosed on point-prevalence rate of 5% is comparable admission. Thirdly, we used clinician-based with other studies, including a recent Centers diagnosis criteria for respiratory tract infec- for Disease Control and Prevention (CDC) tions to reflect ‘real-world’ practice, since study of acute care hospitals in the US, which the majority of these patients are prescribed showed that on any given day, about 4% antibiotics. This possibly over-estimated patients had at least one healthcare-related the true prevalence of infections. Fourthly, infection.5 Eurosurveillence data from acute our assessment of cumulative HAI prev- care hospitals in Europe between 2011 and alence was based on hospital stay till the 2012, showed that on any given day 5.7% of study day, which would not capture patients patients (ie, one in 18 patients) had at least with HAI during the remaining period of one HAI.6 In our study population, UTIs hospitalisation. Despite these limitations, were the most common HAIs and were often we highlight the continuing high burden of catheter-associated. Two-thirds of those with infectious diseases in adult patients admitted catheter-related UTIs had long-term urinary to hospital with consequent pressure on catheters and in these patients hospital onset bed numbers, as well as antibiotic usage. In infections would not be easily preventable. In addition, our cumulative rate of hospital-ac- this study, the mean duration of hospital stay quired infections of approximately 10% in patients with HAI was 33 days, compared has major implications in terms of costs to to the non-HAI patients (15.1 days), and the both healthcare institutions and patients. duration of stay prior to HAI diagnosis (14.5 Risk factors for HAIs are well known and days). Since this study was not specifically include devices such as intravenous lines, designed to assess the contribution of HAI to indwelling catheters, broad-spectrum anti- length of hospital stay, we are unable to draw biotics (predisposing to C.difficile) etc. Some any conclusions from this observation. factors are modifiable and best practice HAIs are costly, both financially for the recommends avoiding unnecessary devices, healthcare facility and for the patient in removal of ‘idle’ catheters, strict adherence terms of loss of income and effects on of hand hygiene, and rationalisation of quality of life. A study by Burns et al in antibiotic use to reduce the risk of these 2010 looked at the cost of HAIs, and in infections. particular the cost of blood stream infec- In conclusion, this study shows that infections occurring in adult patients admitted to tions continue to be a significant source of Auckland City Hospital.7 They showed that morbidity in hospitalised adult patients in the cost of hospital-associated blood stream New Zealand, both at the time of admission infections (HA-BSIs) in a combined group and during hospitalisation.

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 73 www.nzma.org.nz/journal ARTICLE

Competing interests: Nil Author information: Kerry Read, Infectious Diseases, North Shore Hospital, Auckland; Hasan Bhally, Medicine and Infectious Diseases, Waitemata District Health Board, Auckland. Corresponding author: Hasan Bhally, Medicine and Infectious Diseases, Waitemata District Health Board, Auckland [email protected] URL: www.nzma.org.nz/journal/read-the-journal/all-issues/2010-2019/2015/vol-128-no-1426-4- december-2015/6748

REFERENCES: 1. Baker MG, Barnard LT, Morris AJ. Modelling N Engl J Med. 2014 Mar Kvalsvig A, Verrall A et al. the costs of hospital-ac- 27; 370(13):1198-208 Increasing incidence of quired infections in New 6. European Centre for serious infectious diseases Zealand. Infect Control Disease Prevention and and inequalities in New Hosp Epidemiol. 2003 Control. Point prevalence Zealand: a national epide- Mar; 24(3):214-23. survey of healthcare-as- miological study. Lancet 4. Management of Hospi- sociated infections and 2012 Mar 24; 379:1112-19. tal-acquired Infections. antimicrobial use in Euro- 2. Graves N, Nicholls TM, Report of the Controller pean acute care hospitals. Wong CG, Morris AJ. The and Auditor General, Stockholm: ECDC; 2013. prevalence and estimates New Zealand, Volume of the cumulative inci- 1 of 2. http://www.oag. 7. Burns A, Bowers L, Pak dence of hospital-acquired govt.nz/2003/hospi- NT, Wignall J et al. The infections among patients tal- infections/docs/ excess cost associated with admitted to Auckland hospital-vol-1.pdf healthcare- associated bloodstream infections at District Health Board 5. Magill SS, Edwards JR, Auckland City Hospital. Hospitals in New Zealand. Bamberg W, Beldavs ZG et Infect Control Hosp Epide- al. Multistate point-prev- N Z Med J. 2010 Oct miol. 2003 Jan; 24(1):56-61. alence survey of health 15; 123(1324):17-24. 3. Graves N, Nicholls TM, care-associated infections.

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 74 www.nzma.org.nz/journal ARTICLE

Late-life self-harm in the Waikato region WA de Beer, J Murtagh, G Cheung

ABSTRACT AIMS: Late-life suicide is a growing public health concern in New Zealand. Given that suicide attempt is one of the strongest predictors of future suicide, the aim of this study was to examine the characteristics of older people (aged≥65) who presented to the Waikato Hospital Emergency Department following an episode of self-harm between 1 July, 2010, and 30 June, 2013. METHODS: Existing hospital databases and clinical recording systems for medical and psychiatric records were used to identify the sample. Data was collected retrospectively. RESULTS: Of the 52 cases of elderly self-harm, 63.5% were classified as suicide attempt; 19.2% were self-injurious behaviour with no suicide intent; and 17.3% were self-injurious behaviour where the suicide intent was unknown. Overdose was the most common method (65.4%). 61.5% of the cases reported perceived physical illness as a stressor; while 50% were diagnosed with depression. 13.7% had repeated self-harm in the following 12 months. CONCLUSIONS: This study has highlighted the role of physical illness and depression in older people presenting with self-harm. Routine screening of depression in older people with chronic medical conditions and assertive treatment of depression in primary care should be considered as strategies to reduce self-harm and suicide in older people.

uicide is a major cause of death, both and suicide attempts in older people.6,7 internationally and in New Zealand.1 With advances in health care, the number SIn 2004–06, suicide was the fourth of older people living with chronic medical leading cause of death for Māori males and conditions, and therefore at risk of suicide the second leading cause of death for non- and suicide attempts, will also increase. Māori males in New Zealand.2 The Ministry A history of previous suicide attempt is of Health has developed the New Zealand one of the strongest predictors of future Health Strategy in an effort to reduce the completed suicide, particularly in older rate of suicide and suicide attempts.3 people.6 Older people who previously Late-life suicide has become a growing attempted suicide had a higher mortality public health concern among New Zealand’s rate (from both natural medical causes ageing population. The suicide rate for and completed suicide) than the general older people (65 years and over) was 7.3 population, and they used more lethal per 100,000 people, which is lower than the means.8,9,10,11 A Medline literature review age-standardised rate of 10.6 per 100,000 on attempted suicide between 1985 and people.4 However, the rates for the 65+ age 1994 showed that 9 to 18% of older people group ranged widely, from 5.7 (age 65–69) who had made a suicide attempt would to 22.2 (age 80–84) per 100,000 males, and make further attempt(s) within 12 months.12 from 1.3 (age 70–74) to 10.5 (age 85+) per Of the 101 suicide attempters in another 100,000 females. The absolute number of suicide and suicide attempts among older study, two people completed suicide within people are likely to rise as the proportion of one month, while six people made further 9 older people in the population increases.5 non-fatal attempts within 12 months. Chronic medical conditions have been iden- A recent retrospective study conducted in tified as a significant risk factor for suicide New Zealand found that almost a quarter

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 75 www.nzma.org.nz/journal ARTICLE

(24.4%) of older people who committed aged 65 years and older.14 Two databases suicide had a past history of suicide attempt; were accessed to identify episodes of this association was highest (51.6%) in self-harm: 13 females aged between 65–79 years old. A 1. The Consultation-Liaison (CL) French study also showed that older women psychiatry service referral database were more likely to be involved in repeated was used to identify older people suicide attempts.8 Mood disorder was who had been referred for psychi- commonly associated with late-life suicide atric assessment by the ED following attempt, with over two-thirds of older people self-harm. who attempted suicide receiving a diagnosis 2. The existing electronic coding reports of depression.8,9,11 used by the ED and hospital medical The literature on late-life suicide and records department allowed the identi- suicide attempts is, however, limited in New fication of older people who presented Zealand. One Christchurch case-control with self-harm, but were not reviewed study examined 53 adults aged 55 and older by the CL service eg, after-hours who died by suicide or made medically admission (who were either assessed serious suicide attempts, and the risk of by on-call staff or the Crisis Assessment serious suicidal behaviour was found to be Team), those who were admitted increased among people with current mood directly to the high dependency unit disorders, a history of psychiatric hospital or medical wards for acute medical admissions within the previous year, and intervention. The ICD-10 codes X60-X84 limited social networks.4 for “self-harm”, “suicide attempt” and As elderly emergency department “deliberate self-harm” were used to presentations with self-harm have yet to be identify this group. examined in New Zealand, this study offers Each self-harm presentation was clas- an opportunity to: (i) identify the incidence sified using the Colombia Classification of self-harm presentations to the emergency Algorithm of Suicide Assessment (C-CASA).15 department; (ii) characterise the nature The three suicidal behaviour categories of self-harm behaviour; and (iii) identify used in this study were: (i) suicide attempt; demographic and clinical factors that are (ii) self-injurious behaviour with no suicide associated with self-harm behaviour in intent; and (iii) self-injurious behaviour the Waikato region over a 3-year period. where the suicide intent was unknown. The hope was that a better understanding The following four categories of data were of self-harm behaviour in older people extracted from the medical and psychiatric would inform clinicians, public health records of the subjects identified by the CL practitioners, hospital administrators and psychiatric service and ED databases. policy makers on suicide prevention, risk 1. Socio-demographic factors: Age, assessment and management. gender, ethnicity, marital status, living arrangements. Methods 2. Assessment and diagnosis: Past This was a retrospective, descriptive and current psychiatric diagnosis study. Ethics approval was obtained from (depression, bipolar disorder, schizo- the New Zealand Ministry of Health, Health phrenia), antidepressant prescription & Disability Ethics Committee (Reference: at the time of self-harm, co-existent 14/ST1). physical illnesses (dementia, malig- The study population was older people nancy, terminal illnesses and pain), (age ≥65 years) who presented to the non-psychiatric admission in the past Waikato Hospital Emergency Department 12 months, past history of self-harm. (ED) following an episode of self-harm 3. Information about the self-harm: during the period of 1 July, 2010, to 30 June, Location of self-harm, date of 2013. Of the 20 district health boards in self-harm, method, acute stressors New Zealand, the Waikato District Health (death of first-degree relative, Board is the fifth largest, with a total popu- perceived disability and/or suffering lation of 359,310 and 53,022 people (14.8%) from physical illness, terminal illness

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 76 www.nzma.org.nz/journal ARTICLE

Table 1: Socio-demographic factors and information about the self-harm

Male (N=23) Female (N=29) Total (N=52) n (%) n (%) n (%) Age groups (years) 65–79 17 (73.9) 26 (89.7) 43 (82.7)

≥80 6 (26.1) 3 (10.3) 9 (17.3)

Ethnicity European 20 (87.0) 24 (82.8) 44 (84.6)

Māori 2 (8.7) 1 (3.4) 3 (5.8)

Asian 0 (0) 1 (3.4) 1 (1.9)

Unknown 1 (4.3) 3 (10.3) 4 (7.7)

Marital status Married/de factoa 11 (47.8) 16 (55.2) 27 (51.9)

Other 11 (47.8) 10 (34.5) 21 (40.4)

Unknown 1 (4.3) 3 (10.3) 4 (7.7)

Lived alone No 17 (73.9) 16 (55.2) 33 (63.5)

Yesb 3 (13.0) 9 (31.0) 12 (23.1)

Unknown 3 (13.0) 4 (13.8) 7 (13.5)

Self-harm location Home 13 (56.5) 14 (48.3) 27 (51.9)

Other 1 (4.3) 0 (0) 1 (1.9)

Unknown 9 (39.1) 15 (51.7) 24 (46.2)

Self-harm method Overdose 14 (60.9) 20 (69.0) 34 (65.4)

Laceration 5 (21.7) 3 (10.3) 8 (15.4)

Multiple means 2 (8.7) 5 (17.2) 7 (13.5)

Others (vehicle, CO poisoning, chemical ingestion) 2 (8.7) 1 (3.4) 3 (5.8)

C-CASA Suicide attempt 16 (69.6) 17 (58.6) 33 (63.5)

Self-injuries behavior with no suicide intent 6 (26.1) 4 (13.8) 10 (19.2)

Self-injuries behavior with unknown suicide intent 1 (4.3) 8 (27.6) 9 (17.3)

a 2013 census: 62.1% older people (65+) were in partnership (spouse/de facto/partnered)29 b 2013 census: 28.8% older people (65+) were in a one-person household29

NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 77 www.nzma.org.nz/journal ARTICLE

Table 2: History of psychiatric & physical conditions and follow-up in 12 months

Male (N=23) Female (N=29) Total (N=52) n (%) n (%) n (%) Already under mental health service Yes 7 (30.4) 13 (44.8) 20 (38.5)

No 16 (69.6) 16 (55.2) 32 (61.5)

Depression Yes 10 (43.5) 16 (55.2) 26 (50.0)

Depressive Symptoms 5 (21.7) 2 (6.9) 7 (13.5)

No 8 (34.8) 11 (37.9) 19 (36.5)

Schizophrenia/Schizoaffective disorder/Psychosis NOSa Yes 7 (30.4) 0 (0) 7 (17.3)

No 16 (69.6) 0 (0) 16 (82.7)

Bipolar disorder Yes 0(0) 2 (6.9) 2 (3.8)

No 0(0) 27 (93.2) 27 (51.9)

Dementia Yes 4 (17.4) 2 (6.9) 6 (11.5)

No 17 (73.9) 24 (82.8) 41 (78.8)

Unknown 2 (8.6) 3 (10.3) 5 (9.6)

Malignancy Yes 0 (0) 3 (10.3) 3 (5.8)

No 23 (100.0) 26 (89.7) 49 (94.2)

Terminal illness Yes 0 (0) 0 (0) 0 (0)

No 23 (100.0) 27 (93.1) 50 (96.2)

Unknown 0 (0) 2 (6.9) 2 (3.8)

Non-psychiatric hospital admission in past 12 months Yes 17 (73.9) 13 (44.8) 30 (57.7)

No 6 (26.1) 16 (55.2) 22 (42.3)

Past history of self-harm Yes 6 (26.1) 7 (24.1) 13 (25.0)

No 17 (73.9) 21 (72.4) 38 (73.1)

Unknown 0 (0) 1 (3.4) 1 (1.9)

Repeated self-harm in 12 months Yes 4 (17.4) 3 (10.3) 7 (13.7)

No 19 (82.6) 25 (86.2) 44 (86.3)

Unknown 0(0) 1 (3.4) 1 (1.9)

Follow up by mental health service Yes 18 (79.3) 22 (75.9) 40 (76.9)

No 5 (21.7) 7 (24.1) 12 (23.1)

a NOS=not otherwise specified

in first-degree relative or carer, The other stressors that were reported family discord, changed relationship/ by older people: family discord (59.4%), death of friend, relationship sepa- changed relationship/death of friend ration, financial trouble, employment (50.0%), death of first-degree relative change, legal difficulties). (48.5%), financial trouble (37.0%), terminal 4. Longitudinal data 12 months after an illness in first-degree relative or carer index self-harm episode: Repeated (22.9%), relationship separation (20.5%) and self-harm attempt, suicide, deaths. legal difficulties (12.1%). Twenty cases (38.5%) were under the care Results of Mental Health & Addiction Services at the time of self-harm. 50% of the cases had There were a total of 52 self-harm presen- a diagnosis of depression and 14% were tations in the 3-year period. Twenty-three experiencing depressive symptoms. Their (44%) were male (mean age=76 years, background stressors of self-harm were SD=7.9 years) and 29 (54%) were female similar to the entire sample: perceived (mean age=70 years, SD=6.5 years). The illness (60.0%), family discord (40.0%), other demographic data and information about the self-harm are shown in Table 1. changed relationship/death of friend (35.0%), death of first-degree relative Overdose (n=34, 65.4%) was the most (25.0%), terminal illness in first-degree common method of self-harm, followed by relative or carer (15.0%), relationship sepa- lacerations (n=8, 15.4%) and multiple means ration (15.0%) and financial trouble (10.0%). (n=7, 13.5%). Five cases that used multiple means involved an overdose of medication Seven cases (13.7%) repeated self-harm in with alcohol; one case involved an overdose the 12 months following the index self-harm of medication with poisons; and another episode (Table 2). There was no suicide in was an overdose of medication with carbon the 12 months follow-up period, but five monoxide poisoning. Of the ten cases of deaths (cause of deaths unknown in two self-injurious behaviour with no suicide cases) occurred in this period. intent, four cases took an accidental overdose. Histories of psychiatric and physical Discussion conditions are shown in Table 2. Depression The main findings of this descriptive was the most common psychiatric diag- study are consistent with that reported in nosis, 50% of the cases had a diagnosis of the international literature on self-harm/ depression at the time of the self-harm suicide attempt in older people: (i) physical or were given a diagnosis when assessed illness is a significant background stressor; following the self-harm incident. A further (ii) depression is the most commonly diag- seven cases (13.5%) were experiencing nosed psychiatric disorder; and (iii) the rate depressive symptoms, but did not meet of repeated self-harm is high, suggesting the criteria for a diagnosis of depression. this is a very high-risk group.6,7,8,9,11,12 A total of 27 cases (51.9%) were taking Fifty-two cases of older people attempted antidepressant medication at the time of self-harm in the 3-year period. Using self-harm. the 2013 census data on the population The most common background stressor in the Waikato region, this represents a for self-harm was perceived physical illness, 12-month rate of 0.0327% (male=0.0312%; which was found in 32 (61.5 %) cases. female=0.0339%). This rate is slightly higher Examples of physical illnesses included than the 2011 New Zealand national figures chronic obstructive pulmonary disease, on intentional self-harm hospitalisations arthritis and macular degeneration. Pain published by the Ministry of Health (rates was also a common finding at the time of ranged from 0.0198% to 0.0304% for the five self-harm, with 22 (42.3%) cases identi- 5-year age bands aged 65 and older).16 Our fying pain as a factor (in particular, pain study used self-harm emergency department from arthritis and chronic back pain). Over presentations, rather than hospitalisations, half (57.7%) of the cases had at least one as the sampling frame. The higher rate found non-psychiatric hospital admission in the 12 is likely to be explained by the fact that not months prior to the self-harm. every self-harm presentation to an emer-

gency department would result in hospital of age.8,9,11,17,20, In our study, 50% of cases admission. Older females had a slightly fulfilled criteria for a diagnosis of depression higher rate of self-harm than males in our at the time of self-harm, with a further 14% study. Conwell et al suggested that males displaying symptoms of depression without were more successful at completing suicide a definitive diagnosis. Despite this high than females, which may have reduced the prevalence of mental illness, only 39% of number of older males that present to the cases were receiving care from the mental ED.17 Older Māori were under-represented health services at the time of self-harm, in our sample at only 5.8%. The 2013 census highlighting potential detection/screening data recorded 4,690 Māori people aged 65 difficulties for this population group in the years and older living in the Waikato region primary care sector. Suicide prevention and (ie, 11.6% of the region’s total population).14 risk assessment at primary care level may Cheung et al also found an under-repre- provide an opportunity to engage and assess sentation of older Māori in the number of this population, thereby improving clinical completed suicides in New Zealand.13 The management of high-risk individuals. under-representation of older Māori in We found perceived disability and/ suicide and self-harm statistics could suggest or suffering due to physical illness was Māori family and cultural practices may commonly identified as a background protect against elderly suicidal behaviour, stressor of self-harm. A previous case-control but further research is required to explore study found that older people (age ≥65) who this relationship. had attempted suicide had a higher rate of Overdose at home was found to be the medical illnesses as measured by the Cumu- preferred method of self-harm in our study lative Illness Rating Scale.21 Other studies and was consistent with other interna- have also shown an association of physical tional studies.8,11, This might be explained by illness and suicide attempt. For example, the high numbers of prescription medi- Allebeck and Bolund found an increase cations issued to older people (ie, ease of in suicide attempt rate and a diagnosis of access). Overdose on prescribed medica- cancer in men aged 60–69 (SMR=2.3, 95% tions offered a less traumatic and painful CI=1.4–3.5) and aged 80–89 (SMR=2.7, 95% self-harm attempt. In this study, zopiclone CI=1.6–4.5), but not in other groups (men and opiates were the most common medica- aged 70–79; women aged 60–89).22 tions used in overdose. This was in keeping In a 2005 study, the diagnosis of malig- with current literature, which has identified nancy (along with stroke, diabetes mellitus, these medications are frequently prescribed arthritis and bone fracture) was asso- for older people.4,18,19 ciated with an increased risk of attempted After 12 months, 90.4% of cases were alive suicide in older people, but this associ- in this study. We were not able to determine ation was not found in an earlier Japanese the cause of death in two cases, and the study.23,24 Another 2006 study investigated possibility of suicide cannot be excluded. The the life-time history of suicide attempts 12 month repeated self-harm rate of 13.7% and coronary artery disease in people aged found in our study falls within the range of 65 years and older who reported a signif- 9 to 18% reported in a previous literature icant association between suicide attempts review.12 The New Zealand Mental Health and coronary artery disease, even after Survey found community-dwelling older depression was taken into account.25 people (age ≥65) had a 12 month suicide Other studies have suggested that the asso- attempt prevalence of 0.1%. Therefore, our ciation between physical illness and suicide 12 month repeat self-harm rate is over a is seldom direct, but is largely mediated hundred times higher than the general rate in through mood and other mental health the community. Older people with a history of factors.26,27,28 Compared to the international self-harm represent a very high-risk group for literature on physical illness and its rela- repeated self-harm behaviour. tionship to suicidal ideation or suicide, the It has widely been reported in the liter- literature on physical illness and self-harm ature that mood disorders, in particular attempts in older people is limited. depression, are strongly associated with Several limitations to the study design suicide and suicide attempt, regardless have reduced the strength of the current

study. Firstly, although this study reported i. Routine screening of depression and all self-harm cases in a 3-year period, the suicide risk in older people with sample size was small and the sample was chronic medical conditions in primary drawn from one location. Our findings may care and hospital specialist services. therefore not be generalisable to other parts ii. Assertive treatment of depression in of New Zealand. However, the proportion primary care. of older people in our region is similar to iii. Better integration between mental that of the proportion in the New Zealand health services, geriatric medicine, population. Secondly, incomplete medical primary care and hospital specialist records saw a proportion of the data set services for older people with physical missing (reported as “unknown” in the illnesses and depression, particularly tables). In addition, some of the records those with suicide risk. lacked comprehensiveness and speci- iv. Limiting the amount of prescription ficity. Thirdly, the diagnoses of depression, medication to at-risk older people. dementia, bipolar disorder and other v. Active follow-up and treatment for psychotic disorders were based on the older people following an episode mental health clinicians’ clinical impres- of self-harm, including the use of sions, and not on standardised diagnostic evidence-based psychological treatment classifications (eg, DSM, ICD). Further for late-life depression (eg, cognitive prospective studies, with a larger sample behavioral therapy, interpersonal size and including other locations in New psychotherapy and problem solving Zealand, will be useful to examine the therapy) to assist older people to adjust causal relationships with the variables we and adapt to their physical illnesses. have identified in this study. In coming decades, the elderly popu- Despite these limitations, this study has lation in New Zealand will continue to highlighted the importance of perceived increase. More work is needed to address physical illness and depression in older depression-related morbidity and mortality people presenting with self-harm, and this in this vulnerable group. Improving our group of older people represents a very current understanding of late-life suicidal high-risk group of repeated self-harm, and behaviour is required for the development possibly suicide. Along with the inter- of age-specific clinical services and suicide national literature on late-life suicide prevention strategies. Furthermore, iden- behaviour, our findings can be used to tifying the factors associated with late-life inform a number of intervention points suicidal behaviour can improve the iden- to address self-harm and suicide in older tification of at risk older people and their people. These include: clinical management.

Competing interests: Nil Acknowledgements: The authors would like to thank the Waikato District Health Board & the Waikato Clinical School (University of Auckland) for awarding Jonathon Murtagh a summer studentship in 2014/2015. Author information: Wayne de Beer, Specialist Psychiatrist & Clinical Training Director, Mental Health & Addiction Services, Waikato District Health Board, Hamilton, New Zealand; Jonathon Murtagh, Summer Studentship Programme, The Waikato Clinical Campus (The University of Auckland), Waikato District Health Board, Hamilton, New Zealand; Gary Cheung, Senior Lecturer & Old Age Psychiatrist, Department of Psychological Medicine, Faculty of Medical and Health Sciences, The University of Auckland, Auckland 1142, New Zealand. Corresponding author: Wayne de Beer, Mental Health & Addiction Services, Waikato District Health Board, Private Bag 3200, Hamilton. [email protected] URL: www.nzma.org.nz/journal/read-the-journal/all-issues/2010-2019/2015/vol-128-no-1426-4- december-2015/6749

REFERENCES: 1. Public Health Commission. 11. Kato K, Akama F, Yamada 1996;153:1009–1014. A strategic Direction to K, et al. Frequency 21. Levy TB, Barak Y, Sigler Improve and Protect and clinical features M, Aizenberg D. Suicide the Public Health. of suicide attempts in attempts and burden of 1994;Wellington: Public elderly patients in Japan. physical illness among Health Commission. Psychiatry Clin Neurosci. depressed elderly 2013;67:119–122. 2. Ministry of Health. 2014. inpatients. Arch Gerontol Major causes of death 12. Draper B. Attempted Geriatr. 2011;52:115-117. (all ages). Available from: suicide in old age. Int 22. Allebeck P, Bolund C. http://www.health.govt.nz/ J Geriat Psychiatry. Suicides and suicide nz-health-statistics/health- 1996;11:577–587. attempts in cancer statistics-and-data-sets/ 13. Cheung G, Merry S, patients. Psychol Med. Māori-health-data-and-stats/ Sundram F. Late-life 1991;21:979-984. tatau-kahukura-Māori- suicide and suicide health-chart-book/ prevention in New 23. Tsoh J, Chiu HF, Duber- nga-mana-hauora-tuto- Zealand. 2015;Under stein PR, et al. Attempted hu-health-status-indicators/ editorial review. suicide in elderly Chinese major-causes-death- persons: a multi-group, 14. Statistics New Zealand all-ages. Accessed on controlled study. Am Tatauranga Aotearoa. 13th March 2015 J Geriatr Psychiatry. 2013 Census. Available 2005;13:562-571.; 3. Ministry of Health. The at: http://www.stats.govt. New Zealand Health nz/Census/2013-census. 24. Takahashi Y, Hirasawa Strategy. 2000;Wellington: aspx Accessed on H, Koyama K, et al. Ministry of Health 24th January 2015 Suicide and aging in 4. Ministry of Health. 2014. 15. Posner K, Oquendo Japan: an examination Suicide Facts: Deaths and MA, Gould M, et al. of treated elderly suicide intentional self-harm Classification Algorithm attempters. Int Psycho- hospitalisations 2011. of Suicide Assessment geriatr. 1995;7:239-251. Wellington: Minis- (C-CASA): classification 25. Artero S, AstrucB, try of Health. of suicidal events in the Courtet P, Ritchie K. 5. Beautrais AL. A case FDA’s pediatric suicidal Life-time history of control study of suicide risk analysis of antide- suicide attempts and and attempted suicide in pressants. Am J Psychiatry. coronary artery disease older adults. Suicide Life 2007;164:1035-43. in a community-dwelling Threat Behav. 2002;32:1–9. 16. Ministry of Health. elderly population. Int 6. Conwell Y, Van Orden Suicide Facts: Deaths J Geriatr Psychiatry. K, Caine ED. Suicide in and intentional self- 2006;21:108-112. Older Adults. Psychi- harm hospitalisations 26. Barraclough B. Physical atr Clin North Am. 2011. 2014;Wellington: illness and suicide. 2011;34:451-468. Ministry of Health. In: Suicide, clinical 7. Waern M, Rubenowitz E, 17. Conwell Y, Duberstein and epidemiological Wilhelmson K. Predictors PR, Caine ED. Risk studies. London: Croom of suicide in the old factors for suicide in Helm. 1987:37-46. elderly. Gerontology. later life. Biol Psychiatry. 27. Conwell Y, Thompson 2003;49:328-34. 2002;52:193–204. C. Suicidal behavior in 8. Lebret S, Perret-Vaille E, 18. Le Couteur DG, Hilmer elders. Psychiatr Clin Mulliez A, et al. Elderly SN, Glasgow N, et al. North Am. 2008;3:333-356. suicide attempters: Char- Prescribing in older acteristics and outcome. people. Aust Fam Physi- 28. Garand L, Mitchell AM, Int J Geriatr Psychiatry. cian. 2004;33:777–781. Dietrick A, et al. Suicide in older adults: Nursing 2006;21:1052–1059. 19. Tordoff JM, Bagge ML, assessment of suicide 9. Wiktorsson S, Marlow Gray AR, et al. Medi- risk. Issues Ment Health T, Runeson B, et al. cine-taking practices in Prospective cohort study community-dwelling Nurs. 2006;27:355-370. of suicide attempters aged people aged ≥75 years 29. Statistics New Zealand 70 and above: one-year in New Zealand. Age Tatauranga Aotearoa. outcomes. J Affect Disord. Ageing. 2010;39:574–80. 2013 Census QuickStats 2011;134:333-340. 20. Beautrais AL, Joyce about people aged 65 10. Dennis MS, Wakefield PR, Mulder RT, et and over. Available at: P, Molloy C, et al. A al. Prevalence and http://www.stats.govt. study of self-harm in comorbidity of mental nz/Census/2013-census/ older people: Mental disorders in persons profile-and-summary-re- disorder, social factors making serious suicide ports/quickstats-65-plus. and motives. Aging Ment attempts: A case-control aspx Accessed on 24th Health. 2007;11:520-525. study. Am J Psychiatry. January 2015

Prescriber compliance with renal function monitoring in patients taking dabigatran (Pradaxa) Katie Thorne, Stephen Dee, Sisira Jayathissa

ABSTRACT AIM: To assess whether patients prescribed dabigatran had their renal function monitored in accordance with published guidelines. METHODS: We recruited patients from Hutt Hospital and two large primary care practices if they were prescribed dabigatran between July 2011 and April 2012. We assessed patients prescribed dabigatran for more than a year to ascertain whether renal function was monitored at least annually, in keeping with guidelines. RESULTS: All patients had baseline renal function testing. At baseline, 42 (60%) had an eGFR (estimated Glomerular Filtration Rate) over 60mL/min/1.73m2 and 28 (40%) had eGFR between 30-60mL/min/1.73m2. Median follow up was 46 months. Whilst taking dabigatran, 44 of the 70 patients (63%) had at least annual renal function. CONCLUSIONS: Over one-third of patients taking dabigatran for over a year did not have their renal function monitored in keeping with current guidelines, potentially leading to an increased risk of bleeding. We suggest there is a need for an automated reminder to prompt annual renal function testing.

abigatran is predominantly excreted Renal function is known to deteriorate via the kidneys; up to 80%.1 Renal with advancing age, through both phys- Dimpairment has been shown through iological and structural changes, as well pharmacokinetic modeling to lead to a as increasing frequency of medical condi- prolonged half-life and increased plasma tions such as diabetes, hypertension and concentrations of dabigatran and conse- atherosclerosis.5 These patients are more quently an increased risk of bleeding.2 vulnerable to acute changes in renal Dabigatran has been licensed for use in function, leading to higher plasma levels patients with non-valvular atrial fibrillation of dabigatran and increased bleeding risk. and deep vein thrombosis prophylaxis in It is therefore essential to closely monitor New Zealand since July 2011. Since 2014, it renal function in at-risk older patients and was also licensed for treatment of venous discontinue their dabigatran if there is dete- thromboembolism (VTE).3 rioration in renal function. The national Medsafe data sheet, as well Harper et al detailed a series of cases of as Best Practice Advisory Committee (BPAC) bleeding secondary to dabigatran amongst guidelines, recommend that patients taking patients in New Zealand. They cited the risk dabigatran should have their renal function factors of renal impairment and advanced checked prior to starting dabigatran and age, as well as low body weight, as being monitored at least annually.1,3 Dabigatran significant risk factors for increased risk of should not be prescribed for patients with a bleeding whilst taking dabigatran.6 creatinine clearance of less than 30mL/min. We encountered two patients who had It is also recommended that any patient normal renal function at the time of being with a clinical scenario which may cause an started on dabigatran, but subsequently acute kidney injury, such as severe dehy- presented with deterioration in renal dration, has their renal function checked.4 function and bleeding. One of these patients

Table 1: Baseline patient characteristics

Patient characteristic No. of patients

Age (years) <64 13 (19%)

65–74 22 (31%)

75–80 11 (16%)

>80 24 (34%)

Gender (Female) 38 (54%)

Baseline eGFR (mL/ >60 42 (60%) min/1.73m2)

30–60 28 (40%)

<30 0 (0%)

Weight (kg) <50 1 (1%)

50–100 61 (88%)

>100 8 (11%)

Dose 110mg BD 41 (59%)

AF clinic input Yes 27 (39%) died due to irreversible gastro-intestinal and Hutt Hospital laboratory results, which bleeding, and the other patient survived, are the two laboratories used in this local but required admission to intensive care population. We checked results of the blood and a prolonged hospital admission. tests taken prior to starting dabigatran We could not identify any studies exam- and during the period of time the patient ining adherence to guidelines for renal took dabigatran. ‘Baseline’ renal function function monitoring in patients prescribed was assessed using blood tests taken in the dabigatran. We therefore conducted a month prior to starting dabigatran. Renal study of the previously studied cohort of impairment was determined using serum patients from the Hutt Valley region to creatinine and patients’ age using the estimated glomerular filtration rate (eGFR), examine whether renal function had been (calculated by the Modification of Diet in monitored at least annually, according to Renal Disease formula). published guidelines.7 Data was entered onto Microsoft Excel Methods spread sheet. We present descriptive statistics pertaining to the study questions. We recruited patients from Hutt Hospital Central Regional Ethics Committee and the two largest general practices in granted ethical approval for this study. Lower Hutt city. They were identified by healthcare staff in the hospital and through Primary Healthcare databases. All Results patients must have taken at least one dose In the original cohort, we identified of dabigatran between July 2011 and April 102 patients and recruited 92 patients.7 A 2012.7 For the current study, we excluded further 22 patients were excluded from patients if they had not taken dabigatran analysis for the current study, as they did for at least a year. Registered nurses at not take dabigatran for at least 1 year patients’ Primary Healthcare Organisations and so appropriate annual renal moni- (PHO) were contacted by phone to review toring could not be assessed. Therefore, notes and medication lists to determine the current study population consisted of the period of time that the patient was 70 patients. The indication for dabigatran prescribed dabigatran and confirm the was atrial fibrillation for 68 patients, atrial dose. It was therefore not necessary to flutter for 1 patient and another took it for contact individual patients. recurrent VTE. We assessed renal function monitoring Baseline demographics for the 70 patients by reviewing Aotea (private laboratory who took dabigatran for more than a year provider for the greater Wellington region) are listed in Table 1. Median total follow-up

Table 2: Demographics for groups monitored less than annually and at least annually

Patient Characteristic Monitored less Monitored at than annually least annually (26) (44) Age (years) <64 7 (27%) 6 (13%)

65-74 8 (31%) 14 (32%)

75-80 1 (4%) 10 (23%)

>80 10 (38%) 14 (32%)

Gender F 12 (46%) 26 (59%)

Baseline eGFR (mL/min/1.73m2) >60 17 (65%) 25 (57%)

30-60 9 (35%) 19 (43%)

Weight (kg) <50 0 (0%) 1 (2%)

50-100 22 (85%) 40 (91%)

>100 4 (15%) 3 (7%)

Dose 110mg BD 17 (65%) 24 (55%)

AF clinic 9 (35%) 16 (36%) time from starting dabigatran therapy to the at least once a year, 17 had a baseline eGFR current study was 46 months. The majority over 60mL/min/1.73m2, and 9 of them had of patients included were over the age of 65 a baseline eGFR between 30 and 60mL/ years; median age 77 years (ranging from min/1.73m2 (Table 2). Ten (37%) patients who 34 to 93 years old) and 38 (54%) patients did not have renal function monitoring at were women. 13 (19%) were under the age least once a year were seen at least once in of 65, 22 (31%) were aged 65–74, 11 (16%) the specialist Atrial Fibrillation clinic at the were 75–80 and the remaining 24 (34%) hospital after being started on dabigatran. were over 80 years old. 27 (39%) patients were seen at least once in the Hutt Hospital Discussion Atrial Fibrillation clinic, after being started Our study showed that over a third on dabigatran. of patients included did not have renal Forty-one patients took the 110mg, twice function testing in keeping with published daily dose of dabigatran, and 29 took 150mg guidelines. None of the patients had an twice a day. The reasons for using the lower eGFR of less than 30mL/min/1.73m2 on dose were: age over 75 for 15 patients; baseline testing.1 However, half of the study renal impairment with eGFR below 60ml/ population were aged over 75 years, and min for 13 patients; and both age and renal it is recognised that renal function is more impairment for 5 patients. The rationale likely to deteriorate in elderly patients. for using the lower dose in the remaining 8 More than a third of the patients in the patients is unclear as they were under the group who did not have at least annual age of 75 years and all had an eGFR over renal function monitoring were seen in a 60mL/min/1.73m2. Prior to starting dabig- hospital Atrial Fibrillation clinic at least atran, 28 (40%) had an eGFR under 60mL/ once after being started on dabigatran min/1.73m2and none had an eGFR under therapy. This would suggest that there is 30mL/min/1.73m2. a need for increased vigilance for those At the start of treatment, 42 patients had working in both secondary and primary a baseline eGFR of over 60mL/min/1.73m2. care centres. Increasing involvement Whilst taking Dabigatran, 44 (63%) had of specialist reviews may not increase their renal function monitored at least adherence with recommended renal annually. Just over half (57%) of the patients function monitoring. in the group who had at least annual renal It appears that prescribers of dabig- function testing had a baseline eGFR above atran are aware of the need for assessing 60mL/min/1.73m2. In the group of 26 patients renal function monitoring prior to starting who did not have renal function checked therapy, since all patients had a baseline

blood test to assess renal function prior however, it could be argued that there to starting dabigatran. However, our was a large educational campaign to help study suggests that there is not adequate support prescribers and so there should knowledge of, or adherence to, the guide- have been adequate knowledge about lines around need for ongoing renal dabigatran and its use, including the need function monitoring. Other possible for ongoing renal function monitoring. contributing factors for infrequent renal Estimated glomerular filtration rate and function monitoring could include busy not creatinine clearance was recorded in workload, repeat prescriptions or scripts this study due to not having access to the being issued by locum doctors, lack of a patients’ weight for necessary calculations. prompt in electronic prescribing systems We did not differentiate between blood as well as patient factors such as not tests that were completed as part of routine getting the blood tests completed. care and those which were due to a clinical scenario, where renal function change may Although now somewhat controversial, be seen. There was also limited information dabigatran was publicised at its intro- gathered regarding patients’ other medica- duction as having “No need for regular tions and medical conditions. blood tests to see if blood-thinning level The results from this study may not be is in the right range”.9 This view has been directly applicable to others, given that it subsequently challenged. It may have led to was conducted in a small centre in New a perception that regular blood tests are not Zealand. This population does, however, required and so less patient and physician represent a real world group of mostly recognition of the need for regular blood older patients, many of whom have renal tests to check renal function monitoring. impairment and raised important issues An intervention which may improve that are relevant to many clinicians. frequency of renal function monitoring for patients taking dabigatran would be Conclusions an electronic alert every 6 or 12 months, Over one-third of patients included in this similar to those used already for tests such study who took dabigatran for over a year as cervical smear screening. did not have their renal function monitored The numbers included in this study were at least annually. We believe there is a need too small to look at any adverse outcomes for increased education and promotion of to assess whether patients who did not have awareness around not only assessing renal renal function closely monitored were more function when initiating dabigatran, but likely to have adverse events. It is, however, also continuing to monitor appropriately. well recognised that patients who are There may also be a role for an electronic taking dabigatran and develop worsening prompt to be implemented in primary renal impairment are at an increased risk care. We have already been in contact of bleeding.6 with Primary Health Care Organisations in Limitations of this study include the the Hutt Valley region to introduce these small number of patients, due to a large measures, and if successful hope to expand proportion excluded because they were to other regions. prescribed dabigatran for less than a year. In addition, it is a reminder to all Patient recruitment for the original study healthcare practitioners to remain vigilant was also somewhat opportunistic and it is and check patients’ renal function if possible some patients were missed, though they encounter any circumstances in this is felt less likely since several patients which deterioration in renal function were identified by both primary and could occur. If worsening renal function secondary care services, requiring duplicate is identified, dabigatran dose should be entries to be removed. It is also a sample appropriately reviewed regarding whether of patients who were initiated on dabig- dose reduction or alternative medication atran shortly after it became licensed for should be prescribed. use in New Zealand, therefore prescribers We hope our study will help further may have been less familiar with the drug educate and act as a reminder of the need and required monitoring. Alternatively for regular renal function monitoring, as

well as to support development of appro- further information on safety of prescribing priate safety systems into prescribing dabigatran and adverse events in those dabigatran. Larger national and interna- patients who do not receive appropriate tional data-based studies may help to add renal function monitoring.

Competing interests: Nil Author information: Katie Thorne, Hutt Hospital, Wellington; Stephen Dee, Medical Department Hutt Valley DHB, Wellington; Sisira Jayathissa, Medical Department, Hutt Valley District Health Board, Wellington Corresponding author: Katie Thorne, Hutt Hospital, Wellington [email protected] URL: www.nzma.org.nz/journal/read-the-journal/all-issues/2010-2019/2015/vol-128-no-1426-4- december-2015/6750

REFERENCES: 1. http://www.medsafe.govt. 4. BPAC guideline. Renal J Med 2012; 366: 864–6 nz/profs/datasheet/p/ function testing in 7. Thorne K, Jayathissa Pradaxacap.pdf people taking dabigatran S, Dee S, et al. (2014) 2. Chin PKL, et al. (2014). (2012). Accessed 3rd Adherence and outcomes “Correlation between of November 2014 of patients prescribed trough plasma dabiga- 5. Glassock RJ, Rule AD. The Dabigatran (Pradaxa) in tran concentrations and implications of anatomical routine clinical prac- estimates of glomerular and functional changes of tice. Internal Medical filtration rate based the aging kidney: with an Journal, 44: 261–265. on creatinine and emphasis on the glomer- doi: 0.1111/imj.12370 cystatin C.” Drugs in R uli. Kidney International 8. Berthelot E, et al. (2014). and D 14(2): 113-123 82, 270–277; doi:10.1038/ “Impaired renal function 3. Best Practice Advocacy ki.2012.65; published and bleeding in elderly Centre (BPAC). The use online 21 March 2012 treated with dabigatran.” of dabigatran in general 6. Harper P, Young L, Blood Coagulation and practice: a cautious Merriman E. Bleeding Fibrinolysis 25(6): 618-620. approach is recommend- risk with dabigatran in 9. https://www.pradaxa.com/ ed. BPJ. 2011; 38: 10–27. the frail elderly. N Engl pradaxa-vs-warfarin

What makes a child a ‘competent’ child? Amanda van Rooyen, Tineke Water, Shayne Rasmussen, Kate Diesfeld

ABSTRACT Competence is a vital component of the informed consent process. The perceived level of a child’s competence may influence their degree of participation in health decisions that affect them. It is the responsibility of the health professional to gauge a child’s level of competence. Child competence, however, is not a static attribute that is linked to age. Rather, it is dynamic, changing in nature and dependent on a child’s previous experiences, personal attributes, network of relationships around them and cultural and environmental context. Consequently, there is no single verified assessment tool to assist in the recognition of competence for New Zealand children. Adding to this complexity are the unclear interpretations of New Zealand health legislation and policy regarding whether or not a child can legally consent or refuse healthcare advice and treatment without the consent of a legal guardian. Under the Care of Children Act 2004 and the Code of Health and Disability Services Consumers’ Rights 1996, the Health and Disability Commissioner1 states “a child may consent themselves [to health treatment] if and when the child achieves sufficient understanding and maturity to understand fully what is proposed”. This paper poses the question: What is ‘competency’ and how is this decided? For the purpose of this article, ‘child’ pertains to those under the age of 16 years.

he threshold for child competence to informed consent are effective commu- consent to health care treatment in nication of full information, based on the TNew Zealand remains ambiguous. Al- voluntary choice of a competent person.3 To though not clearly stated, New Zealand law make competent decisions, children require infers that children under the age of 16 years age appropriate information that supports may give or withhold consent to healthcare their health literacy and the opportunity treatment, so long as they are competent to to participate.4 Children’s competence is do so.1,2 Currently, it is the role of the health- defined by Alderson 5 as “more than a skill, care professional to decide whether or not it is a way of relating and can be under- a child has adequate competence, which ac- stood more clearly when each child’s cording to the Health and Disability Commis- inner qualities are seen within a network sioner, is the time at which a child achieves of relationships and cultural influences”. sufficient understanding and maturity to Recognising competent children not only fully comprehend the proposed treatment.1 supports ethical arguments regarding Upholding a child’s decision through the respect for children’s rights and their informed consent process is one way of en- personhood; it has other more tangible suring children’s rights to participation and benefits to both the child and healthcare autonomy are respected. However, compe- services. These include improved treatment tence is not only a legal concept, it is also the adherence, clinical effectiveness, disease degree to which health professionals allow prevention and delivery of health children to participate in matters that are services.6-9 Secondary benefits include important to them. children learning to advocate and take The process of informed consent, as responsibility for their own health, and explained by the Ministry of Health,3 obli- the enhancement of their personal devel- 8,10 gates health professionals to acknowledge opment and participation in society. and respect a health consumer’s right to The multiple benefits of involving and autonomy. The main principles underlying respecting competent children in decision

Figure 1: Internal and external influences on children’s competence to give informed consent.5,24

making may be evident, however the act of organisations and representatives from identifying competent children is complex two Independent Crown entities, the Office and fraught with difficulties. First and of the Children’s Commissioner and the foremost, competence is a dynamic state Human Rights Commission. The report rather than a fixed attribute that a child ‘Kids Missing Out’,19 released by UNICEF in either does or does not possess. December, 2013, was a stocktake of New There are multiple internal and external Zealand’s progress on implementing United factors that influence the state of a child’s Nations Convention on the Rights of the competence. Some internal factors include Child (UNCRC). It stated: prior experiences of illness,11 level of inde- Initiatives to allow children to have a pendence, ethnicity and temperament.12 say in matters that affect them have External factors include the environment not always been sustained, there are in which competence is assessed12 and very few processes for eliciting chil- the manner and form in which infor- dren’s views on legislative and policy mation is imparted.3,13-18 Changes in social development, and children’s ability and cultural contexts,19 media represen- to participate in judicial and admin- tation, and family and health professional istrative proceedings is variable. support can also have an impact on Child participation is also recognised a child’s level of competence.4,5,8,11,12,20 by the Human Rights Commission in the Culture may dictate the manner by which National Plan of Action for the Promotion health professionals impart information and Protection of Human Rights (NPA),22 and conduct the informed consent which was due to be completed in June process. The interface between different 2015. Taking into account the opinions of cultural worldviews of health profes- children in the Children’s Symposium 2003, sionals, children, and the children’s family the NPA acknowledged that children need may bring about opportunities for misun- to be listened to, have their opinions given derstandings. On a practical basis, Brook due weight and have their participation

(2000)16 acknowledges that competence is rights under UNCRC implemented.23 The difficult to define, assess and measure. NPA suggested the creation of a program Recommendations for improving New to improve children’s participation in Zealand children’s participation in society governmental and non- governmental and health decisions were given by the UN sectors, including educational resources to assist organisations to involve children in Committee on the Rights of the Child (CRC) decision making.23 in 1997, however the UNCRC Monitoring Group determined that the level of children’s participation has not progressed.21 What is ‘competency’? The UNCRC Monitoring Group has repre- Children’s competence is not a fixed state sentatives from non-governmental but is dynamic; their ability to understand

develops and modifies with their experi- but to kaumātua and kuia (grandparents) ences and changes in their social contexts.12 and tīpuna (ancestors); they are a part of In addition, children’s competency may a whānau, hapū (extended family) and be recognised, denied, encouraged or iwi (tribe), and they belong to the whenua inhibited.11,12 Whether children’s choices (land).3 Consequently, when a child is are honoured may be dependent upon the faced with a health decision, so may be supporting adults’ willingness to support- the whānau and hapū, rather than the ively, generously and courageously respect simple dyad of parent and child, which is children’s decisions.11,12 Figure 1, described frequently the focus of Anglo-European by Alderson (1992)5 and illustrated by approaches. A Māori child’s competence Orr (1999),24 aptly depicts the myriad of may be influenced if they are approached internal and external variants that may on individual terms rather than receiving influence a child’s competency. Although it support in a more collective manner.3 The fails to include facets specific to the health collective approach may also be applicable care environment, such as life experience, to many Pacific children. Pacific families nature of information and its delivery, come from 22 different countries, all and opportunity to participate, it provides varying in their use of the English language, a concise description of intrinsic and and their involvement in church and extrinsic factors that may affect a child’s other supporting groups.3 Many Māori and level of competency. Pacific children are part of a large extended Age has been shown to be an inaccurate family and community; their wellbeing is marker of the level of children’s compe- contingent on their integration and on the 3 tence.12,25 Whereas children’s experiences, community’s overall wellbeing. However, both in general life adversities and in it is important to also acknowledge that illness, have been found to more greatly diversity exists within the collectivism influence their capacity to give informed approach for many contemporary Māori consent.11,12 The ability of children to and Pacific Island families. The Charter develop health literacy and demonstrate on the Rights of Tamariki Children and competence is impacted by additional Rangatahi Young People in Healthcare 26 factors such as the use of technical language Services in Aotearoa New Zealand and the pace at which information is supports the rights as set out by UNCRC and imparted. It is the responsibility of health that children’s health status based on tino professionals to impart information in a rangatiratanga and te Tiriti o Waitangi are way that supports children developing a vital ingredient for the provision of health health literacy. This influences a child’s services to Māori children. A child’s culture ability to process and understand their and local ecosystem need to be taken into conditions and options. In addition, the consideration by health professionals who environment in which the information is are attempting to understand a child’s 27 given may also have an effect, for example experiences, capabilities and perspectives. the unfamiliar hospital ward environment Kawa whakaruruhau (cultural safety) is an versus the community setting. The time important aspect of these considerations, a child has to digest and understand the which require an understanding and accep- information is another relevant factor (and tance of cultural differences. may be a barrier to obtaining meaningful consent in an acute setting). Why is it important Different cultural constructions of to recognise childhood, the family, and healthcare settings may influence the manner and competent children? form of information imparted, as well as It has been observed that respecting the manner by which it is interpreted.3 For children’s involvement in health care example, the Anglo-European emphasis decision- making contributes to the on individual autonomy may conflict with improvement in their health status.3 Doyle, the Māori value of wholeness and collec- Lennox, and Bell9 presented this view in tivity. Many Māori tamariki (children) their systematic review, which positively may not only belong to mātua (parents) associated patients’ experiences with

patient safety and clinical effectiveness. Involvement in healthcare decisions and What is known respect for patients’ preferences were two about New Zealand of four relational aspects used to measure patients’ experiences.9 A positive association health law and policy was then found between patients’ expe- with regard to child riences and adherence to recommended medication and treatments, preventative competency? care, healthcare resource use and tech- Children’s competence to consent in New nical quality-of-care delivery.9 Coyne and Zealand is regulated by the Code of Health Gallagher utilised similar definitions for and Disability Services Consumers’ Rights 1996 (the Code)29 and the Care of Children children’s experiences in healthcare settings. Act 2004.30 It is also influenced by foreign They identified that children who were case law, most notably the Gillick case,31 involved in the decision making process and also the UNCRC32, which introduced had mostly positive experiences and that the notion of diminishing parental respon- the process helped them prepare for what sibility with the evolving capacity of the to expect, reduced their anxieties, and child. The Ministry of Health3 discussed provided reassurance.8 Consequently, the this notion as the ‘maturity approach’, recognition of children’s capacities and which may be contrasted with the ‘status their involvement in decision making is approach’ in which the age of the child is an integral part of their healthcare expe- determinative.30 rience, and in turn patient safety and clinical effectiveness. The 1985 landmark English case of Gillick vs West Norfolk and Wisbech Children’s views of their healthcare AHA31 was significant because it marked experiences provides vital information.7 the emergence of the ‘competent child’ There is a dearth of research regarding discourse. It recognised that children can be their views of health services and existing independent, autonomous, and competent research is often from adult proxy deci- decision makers with regards to their 7 sion-makers. In particular, the views health care. It challenged the ideology of of younger children were rarely sought, parenting being a right, or a dominant despite growing evidence of their compe- and controlling process in a child’s life, to tence to provide valuable contributions to being a responsibility and duty.33 This case 7 healthcare service improvement. Health is a clear rejection of the ‘status based’3 professionals may gain a true represen- approach, where a set age limit dictates tation of children’s needs by directly the competency of a person.17 The House seeking their feedback. of Lords ruled that parental rights to Listening to children and respecting decide whether or not their child receives their opinions can contribute to a child’s medical treatment cease when the child personal development. This support reaches sufficient maturity and has the can lead to children making better deci- understanding and intelligence to make sions which can lead to improved health an informed decision.34 The Ministry of outcomes.10 It may prepare them to Health,3 Medical Council of New Zealand35 participate in society and strengthen their and Health and Disability Commissioner1 accountability.10 Allowing children to have all indicated their growing support for the an active role in their healthcare decisions maturity based approach,3 which supports teaches them in an incremental process the applicability of the Gillick case to the rather than having instantaneous respon- New Zealand context. sibility at the age of sixteen.28 Treatment Unfortunately, there has been little is more likely to be effective if children guidance as to what Gillick competence are allowed to take part in the decision actually is, and when or how it can be making and for their contributions to be applied in New Zealand.32 This leaves a respected; alternatively, children who feel grey and ill- defined area for New Zealand coerced into medical treatment tend to courts, and an ethically challenging set of recover more slowly.6,12,28 principles for health professionals.32

United Nations Convention on the theless, the better view is that minors’ Rights of the Child common law capacity to consent to medical treatment has not been The seminal document UNCRC,32 ratified extinguished by the New Zealand by New Zealand in 1993, obligates health legislation, and that the consent of professionals to ensure children’s voices those under the age of 16 will some- are heard and given due weight in accor- times be effective in law, be it for the dance with their level of maturity. However, purpose of the criminal law, the law the notion of ‘the child’s best interests’ of torts, or the Code of Rights. described in Article 3 of UNCRC overrides children’s rights of self- determination, Thus, New Zealand awaits further freedom of expression (Article 13), and detailed legal guidance. respect for their views (Article 12).32 The Medical Council of New Zealand The Code of Health and Disability The Medical Council of New Zealand Services Consumers’ Rights 1996 (MCNZ) supports the assessment of a child’s competency to give informed consent. New Zealand is unique in that the Code The MCNZ’s guidance, contained in Infor- presumes all consumers of healthcare to mation, Choice of Treatment and Informed be competent.17 Right 7(2) states, “Every Consent 2011, described a competent child consumer must be presumed competent as an individual who “is able to understand to make an informed choice and give the nature, purpose and possible conse- informed consent, unless there are quences of the proposed investigation or reasonable grounds for believing that the treatment, as well as the consequences consumer is not competent”.29 Right 7(7) of non-treatment”.35 However, at present addresses refusal of consent by stating, there is a paucity of research evidencing “Every consumer has the right to refuse how health professionals put this advice services and to withdraw consent to into practice and conduct their assessment services”.29 ‘Consumer’ has been defined of child competence to consent. The MCNZ as a health or disability services consumer recognised the lack of direction on the and does not exclude children. Although subject from the Care of Children Act 2004.35 not directly supporting the maturity-based It stated: approach, the presumption of competence It is not clear whether parental rejects the status-based approach. consent is always necessary for Care of Children Act 2004 medical treatment or procedures for In contrast to the Code, the Care of persons under 16 years. Section 36 Children Act 2004 (s36) refers to age—a [of COCA] does not automatically status-based approach.3,35 The Act is less prohibit persons under 16 years from clear regarding people under 16 years of consenting to medical, surgical or age compared to those over 16 years of age. dental procedures. In the absence of This is despite there being a large number clear legislative direction it is likely of submissions on the Care of Children that the principles laid out in Gillick, Bill recommending clarification on the namely that parental consent is not issue and proposing adoption of the rule of always necessary for medical proce- ‘evolving capacities’ in line with the Gillick dures or treatment for persons under case.18 The abstruseness and inconsistencies 16 years, will be followed by New in these governing documents leaves health Zealand courts. professionals with little clarity about how to effectively assess competence and build it Deciding competency into the fabric of interactions with children. Currently, no solitary assessment tool is 2 In 2006, Professor Skegg reported on the applied to assess children’s competence status of consent by competent children in in New Zealand. Rather, it is a judgment New Zealand: made by health professionals based on legal Given the indecisive and conflicting guidance from the Ministry of Health3 and High Court decisions, the matter is the Medical Council of New Zealand.35 The not entirely free from doubt. Never- context in which the healthcare decision

is to be made, for example the acuteness from the MCNZ that state the importance of the child’s illness, the complexity of the of understanding the nature and purpose information and the available time for a of the treatment and its consequences.35 decision to be made, might also influence Lord Scarman and Lord Fraser, from the a healthcare professionals assessment of a Gillick case, stated the need for sufficient child’s level of competency.3 understanding and intelligence.31 Hence, A number of screening tools and frame- healthcare professionals are required to works have been developed in an attempt to make case-by-case judgements on the level standardise the assessment of competence. of perceived competence a child possesses, One of the more recent screening tools which in turn may affect the child’s level of was developed in 1998 by Billick et al,36 involvement in healthcare decisions. who conducted the Competency Question- naire-Child Psychiatric (CQ-ChP) test which Conclusion evaluated 25 inpatient children for compe- Competence is not a rigid dichotomy tency, and utilised the Wechsler Intelligence between competent or incompetent, but Scale for Children-Revised Edition (WISC-R). rather a dynamic continuum. The evolving The main aim of this test was to identify nature of competence makes it difficult an age at which competency was achieved. to state a simple set of rules or attributes The mean age of competency was found a child requires in order to be deemed to be 10.1 years with participants showing competent. The importance of recognising a year 6/7 (10 to 11 years old) reading competent children and giving their views 36 level. However, the authors concluded it due weight may be clear, however the was not possible to correlate competency act of identifying who is competent and 36 with an age. In 2001, the CQ-ChP test was who is not is complex. The inconsistent revised as the Competency Questionnaire- alignment of New Zealand health legislation Pediatric Outpatient Modified Version and policies further obscures this process. (CQ-Peds), which consisted of 19 items and However, the UNCRC and the Charter chal- emphasised the developmental aspect of lenge us to seek greater participation and 37 competence in children. Again, an age at decision power sharing with children. which competency was achieved could not This article highlights the need for further be determined.37 action from both academic and govern- The document, Consent in Child and mental agencies to address the issues faced Youth Health: Information for Practitioners in determining the competence of children by the Ministry of Health, 1998,3 indicated to make health decisions. An exploration the provisions for a child to be deemed of the key attributes of child competence is competent. It stated: required to assist health professionals in the Regardless of age, to be deemed identification of competent children, which competent an individual must be should inform more clear and practical able to understand that they have policies on the subject. This paper provides a choice (freedom from coercion), the basis for further research by the authors why they are being offered the to include a definition of child competence ‘treatment’, what is involved in what relevant to the New Zealand health context, they are being offered, and what the a tool to assist health professionals to probable benefits, risks, side effects, identify competence attributes in children failure rates and alternatives are. and professional development programs for Although it pre-dates the Care of Children health professionals to support the partici- Act 2004, it is consistent with instructions pation of children in health care.

Competing interests: Nil Acknowledgements: Authors would like to acknowledge Amanda B Lees for her review of this work. Author information: Amanda van Rooyen, Clinical Educator and MPhil Candidate School of Clinical Sciences, Auckland University of Technology, Auckland; Dr Tineke Water, Research Leader, School of Clinical Sciences, Auckland University of Technology Nursing Research Fellow, Starship Children’s Health, Auckland; Shayne Rasmussen, Associate Head of Postgraduate Studies and Lecturer, School of Clinical Sciences, Auckland University of Technology Auckland; Professor Kate Diesfeld, Professor, Health Law Department of Public Health, Auckland University of Technology, Auckland Corresponding author: Amanda van Rooyen, School of Clinical, Sciences, Auckland University of Technology, AUT North Shore Campus, Private Bag 92006, Auckland, 1142, New Zealand [email protected] URL: www.nzma.org.nz/journal/read-the-journal/all-issues/2010-2019/2015/vol-128-no-1426-4- december-2015/6751

REFERENCES: 1. Health and Disability Water T. Children’s voices children? Minors’ consent Commissioner. Fact Sheet in public hospital health- to health care treatment 3: The age of consent and care delivery. NZ Med J. and research. Soc Sci Med. informed consent for 2014, 127(1405), 24-31 2007, 65(11), 2272-2283 children. [Online] 2014. 8. Coyne I, Gallagher 13. Bluebond-Langner M. http://www.hdc.org.nz/ P. Participation in The private worlds publications/other-publica- communication and of dying children. tions-from-hdc/fact-sheets decision-making: Chil- Princeton University (accessed Jan 13, 2014) dren and young people’s Press: NJ, USA, 1978. 2. Skegg PDG, in Medical experiences in a hospital 14. Christensen PH. Difference law in New Zealand; setting. J. Clin Nurs. and similarity: How Skegg PDG, Paterson 2011, 20, 2334-2343. children’s competence is R, Eds.; Brookers Ltd: DOI: 10.1111/j.1365- constituted in illness and Wellington, New Zealand, 2702.2010.03582.x its treatment. In Children 2006; pp 191-195. 9. Doyle C, Lennox L, Bell and social competence: 3. Ministry of Health. D. A systematic review Arenas of action; Hutchby Consent in child and youth of evidence on the I, Moran-Ellis J. Eds. The health: Information for links between patient Falmer Press: London, practitioners. [Online] experience and clinical UK, 1998; pp. 187-201. 1998. www.moh.govt.nz safety and effectiveness. 15. Coyne I. Consultation (accessed Jan 13, 2014) Br Med J Open. 2013, with children in hospital: 4. Martenson EK, Fagerskiold 3(1), 1-18. DOI:10.1136/ Children, parents’, and AM. A review of children’s bmjopen-2012- 001570 nurses’ perspectives. J. decision-making compe- 10. Lansdown G. Every Clin Nurs. 2006, 15, 61-71. tence in health care. J. Clin child’s right to be heard: 16. Brook G. Children’s Nurs. 2008, 17, 3131-3141. A resource guide on the competency to consent: A DOI: 10.1111/j.1365- UN Committee on the framework for practise. 2702.2006.01920.x Rights of the Child general 17. Paediatr Nurs. 5. Alderson P. In the genes comment No.12.; Save the 2000, 12, 31-35. or in the stars. J. Med Children UK: London, 2011 18. McLean K. Children and Ethics. 1992, 18, 119-124. 11. Alderson P, Sutcliffe competence to consent: 6. Alderson P. Consent K, Curtis K. Children’s Gillick guiding medical to children’s surgery competence to consent treatment in New and intensive medical to medical treatment. Zealand. VUWLR. 2000, 31, treatment. J. Law & Hastings Cent Rep. 551-576. Miller, F. Wake Soc. 1990, 52, 58. 2006, 36(6), 25–34 up COCA! Give children 7. Dickinson A, Wrapson W, 12. Alderson P. Competent the right to consent

to medical treatment. right to consent. J. Clin 31. www.legislation.govt.nz NZFLJ. 2011, 7(4), 1-9. Nurs. 1999, 8, 291-298. (accessed Jan 13, 2015) 19. UNICEF New Zealand. 25. Weithorn L, Campbell 32. House of Lords. Gillick Kids missing out. [Online] S. The competency of v Wisbech and West 2013. www.unicef.org.nz children and adoles- Norfolk Area Health (accessed Jan 15, 2015) cents to make informed Authority. [Online] 1985. 20. Curtis K, Liabo K, Roberts treatment decisions. Child http://www.hrcr.org/ H, et al. Consulted but Dev. 1982, 53, 1589. safrica/childrens_rights/ not heard: A qualitative 26. Children’s Hospitals Gillick_WestNorfolk.htm study of young people’s Australasia/Paediatric (accessed Jan 16, 2015) views of their local Society of New Zealand. 33. United Nations General health service. Health Charter on the Rights Assembly. Convention on Expect. 2004, 7, 149-156. of Tamariki Children & the Rights of the Child. 21. UNCROC Monitoring Rangatahi Young People United Nations, Treaty Group. Monitoring the in Healthcare Services in Series, 1577. [Online] implementation of the Aotearoa New Zealand. 1989. www.justice.govt. United Nations Convention [Online] 2010. http:// nz (accessed Jan 16, 2014) of the Rights of the Child www.paediatrics.org.nz/ 34. Grimwood T. Gillick in New Zealand: UNCROC files/Charter%20on%20 and the consent of Monitoring Group final the%20rights%20of% minors: Contraceptive submission. [Online] 20Children%20and%20 advice and treatment in 2013. www.occ.org.nz Young%20People%20 New Zealand. VUWLR. (accessed Jan 15, 2015) in%20NZ%20final%20 2009, 40(4), 743-769. version.p df (accessed 22. Human Rights Commis- 35. Drake C. Review: Informed Jan 20, 2015) sion. New Zealand’s fifth consent? A child’s right to periodic report under 27. Graham A, Powell MA, autonomy. J child health the United Nations Taylor N, et al. Ethical care. 2001, 5(3), 101-104. Convention on the Rights research involving DOI: 10.1177/13674935010 of the Child: A summary. children. [Online] 0500302 [Online] 2015. http://www. 2013. childethics.com 36. Medical Council of New hrc.co.nz/2014/12/19/ (accessed July 1, 2015) Zealand. Information, uncroc-public-consul- 28. Lewis CE, Lewis MA, choice of treatment and tation-have- your-say Lorimer A, et al. Child informed consent. [Online] (accessed Jan 20, 2015) initiates care: The 2011. www.mcnz.org.nz 23. Human Rights Commis- use of school nursing (accessed Jan 13, 2015) sion. The rights of children services by children in 37. Billick SB, Edwards and young people Nga tika an adult free system. JL, Burgert III W, et o nga tamariki, rangatahi. Pediatrics. 1977, 60, 499. al. A clinical study of In National Plan of Action 29. Health and Disability competency in child for the Promotion and Commissioner. Code of psychiatric inpatients. Protection of Human health and disability J. Am. Acad. Psychiatry Rights. [Online] 2015. services consumers’ rights. Law. 1998, 26(4), 587–594. http://www.hrc.co.nz/ [Online] 1996. http:// 38. Billick SB, Burgert III W, report/chapters/chapter04/ www.hdc.org.nz/the-act- Friberg G, et al. A clinical children01.html#tro -code/the- code-of-rights study of competency to (accessed Jan 20, 2015) (accessed Jan 14, 2015) consent to treatment 24. Orr FE. The role of the 30. Care of Children Act. in pediatrics. J. Am. paediatric nurse in Consent to medical Acad. Psychiatry Law. promoting paediatric procedures. [Online] 2004. 2001, 29(3), 298–302.

Maxillofacial fractures at Waikato Hospital, New Zealand: 2004 to 2013 Blake K Moore, Ryan B Smit, Angus N Colquhoun, W Murray Thomson

ABSTRACT Injury to the maxillofacial region continues to place a burden on hospital care in New Zealand, with maxillofacial fractures often being associated with both a significant social cost and personal morbidity. This article describes the characteristics, aetiology and treatment patterns in a tertiary maxillofacial centre in New Zealand during a 10-year period. Over the observation period, a total of 1,975 cases were treated, with a male-to-female ratio of 4:1. The highest incidence was in the 20–29-year-age group. Interpersonal violence (IPV) was the most common aetiology, observed in 54.5% overall, and more common among males than females (58% and 38% respectively; P<0.001). Falls were the most common cause of injury among older females (those aged 50+). Comparison to an earlier analysis shows that IPV-related maxillofacial trauma has increased significantly at this tertiary centre, increasing from 36.2% of cases in 1989–2000, to 54.5% in 2004–2013. There remains an urgent need for appropriate health promotion to reduce interpersonal violence, as well as an increase in the staffing numbers of maxillofacial units in New Zealand.

axillofacial injury continues to A previous analysis of facial fractures in place a burden on the New Zea- the Waikato region found that IPV-related Mland hospital sector. Road traffic injury steadily increased from 31% to 41% accidents (RTA) and interpersonal violence in the first 9 years of their observation (IPV) have been highlighted as the most period.1 1-3 common causes. Maxillofacial fractures International research in Western are often associated with a significant social countries (such as the UK, France and 2 cost and personal morbidity. Finland) also supports a downward trend For some time in New Zealand, RTA-re- in RTA-related fractures, with IPV now lated maxillofacial trauma has reduced and, being the leading cause.9,10 In Denmark, the in some areas, has been less common than incidence of IPV-related fractures doubled falls and sporting injuries.1,4,5 This is mainly during the 1960–1987 period.11 However, due to changes in transport policy, including road traffic accidents remain the leading revised drink-driving laws, reduced road cause of facial fractures in developing speed, modern vehicle safety measures countries such as Brazil,12,13 India and (such as the wider availability of airbags), Iran,14,15 and also in some developed and improved road conditions.5 nations, such as Japan, Greece, and the 16-18 IPV continues to be the leading cause of Netherlands. maxillofacial injury in New Zealand,1-4,6 The multifactorial nature of facial frac- primarily involving young men and with tures means that both their incidence and alcohol as a frequent contributor.1,2,7,8 aetiology vary not only among countries,19 Reports from the Canterbury, Otago, and but also within them. The aim of this study Waikato regions showed that IPV accounted was to describe the characteristics, nature for between 32% and 44% of all facial and treatment of maxillofacial fractures fractures.1,3,4 Anecdotal reports from the presenting to a tertiary referral centre in Waikato region suggest that RTA have New Zealand during a 10-year period, and continued to decrease, while a greater to determine whether trends observed in an proportion of facial fractures are due to IPV. earlier such analysis have continued.

Table 1: Number of maxillofacial fracture cases for the periods 2004–06, 2007– 09 and 2010–13, by sociodemographic characteristics and fracture aetiology (brackets contain column percentages unless otherwise indicated)

Period

2004–2006 2007–2009 2010–2013 Total Sex Male 467 (84.3) 424 (82.2) 714 (79.1) 1,605 (81.3)

Female 87 (15.7) 92 (17.8) 189 (20.9) 368 (18.7)

Age group (years) 1–19 147 (26.5) 142 (27.6) 222 (24.6) 511 (25.9)

20–29 240 (43.3) 193 (37.5) 323 (35.8) 756 (38.4)

30–39 72 (13.0) 73 (14.2) 128 (14.2) 273 (13.9)

40–49 49 (8.8) 52 (10.1) 112 (12.4) 213 (10.8)

50+ 46 (8.3) 54 (10.5) 118 (13.1) 218 (11.1)

18–25 254 (45.8) 198 (38.5) 336 (37.2) 788 (40.0)

Aetiology IPVc 296 (53.4) 293 (56.6) 487 (53.9) 1,076 (54.5)

RTAd 96 (17.3) 69 (13.3) 121 (13.4) 286 (14.5)

Falls 49 (8.8) 58 (11.2) 123 (13.6) 230 (11.6)

Sport 73 (13.2) 65 (12.5) 97 (10.7) 235 (11.9)

Other 39 (7.0) 30 (5.8) 62 (6.9) 131 (6.6)

Unknown 1 (0.2) 3 (0.6) 13 (1.4) 17 (0.9)

All combinede 554 (28.1) 518 (26.2) 903 (45.7) 1,975 (100.0)

aAge data missing for 4 patients; bSex data missing for 2 patients; cInterpersonal violence; dRoad traffic accident;e Row percent

injury, date of injury, and the treatment Method provided. Regrettably, information on The Department of Maxillofacial and ethnicity and alcohol involvement was not Oral Surgery at Waikato Hospital provides collected in the database. Statistical analysis a tertiary service to a regional population was undertaken using SPSS (Statistical of over 846,600 in the upper central North Package for the Social Sciences; SPSS Inc, Island. The catchment areas of service Chicago, Illinois, US; version 20). As far include the regions of Waikato, Taupo, as possible, the analytical and reporting Gisborne, Bay of Plenty, the Coromandel approaches were kept similar to those used Peninsula and Thames. in an earlier analysis of a case series of Clinical data collected from all patients Waikato maxillofacial fractures, in order to with maxillofacial fractures attending enable direct comparisons to be made.1 the Department of Maxillofacial and Oral Following the computation of descriptive Surgery at Waikato Hospital from January, statistics, bivariate associations were 2004, to December, 2013, were retrospec- tested for statistical significance using the tively analysed using information that Chi-squared test. Census numbers for 1991, had been recorded prospectively using 1996, 2001, 2006 and 2013 were used as a paradox database. Patients who had estimates for the number of people at risk sustained facial fractures were included, in each of these years for the Waikato DHB but those with isolated soft tissue injuries catchment population. A linear regression were omitted. Details recorded included model was fitted to give an estimated sex, age, cause of injury, classification of population for each of the years from 1991

Table 2: Rates of fractures by period and aetiology quantified per 100,000 per- Table 3: Percentage of primary reported son-years at risk (brackets contain 95% confidence intervals) cause by gender

Period Gender Aetiology 2004–2006 2007–2009 2010–2013 Aetiology Male (%) Female (%)

IPV 20.7 (CI 18.5–23.2) 20.1 (CI 17.3–22.6) 24.5 (CI 22.4–26.8) IPV 87.2 12.8

RTA 6.7 (CI 5.5–8.2) 4.7 (CI 3.7–6.0) 6.1 (CI 5.1–7.3) RTA 77.2 22.8

Falls 3.4 (CI 2.6–4.5) 4.0 (CI 3.1–5.2) 6.2 (CI 5.2–7.4) Falls 60.4 39.6

Sport 5.1 (CI 4.1–6.4) 4.5 (CI 3.5–5.7) 4.9 (CI 4.0–6.0) Sport 87.2 12.8

Total 38.8 (CI 35.7–42.1) 35.6 (CI 32.6–38.7) 45.5 (CI 42.6–48.5) Other/ 69.2 30.8 Unknown Number of cases 554 518 903

Table 4: Type of maxillofacial fracture cases, by fracture aetiology and treatment (brackets contain column percentages unless otherwise indicated)

Type of maxillofacial fracturea

Mandible Le Fort I Orbit Zygoma Frontal Alveolar Skull Row totals

Aetiology IPVb 614 (61.7) 51 (35.9) 208 (46.3) 230 (49.7) 7 (21.2) 12 (29.3) 58 (43.3) 1,180 (52.3)

RTAc 110 (11.0) 50 (35.2) 93 (20.7) 88 (19.0) 15 (45.5) 4 (9.7) 29 (21.6) 389 (17.2)

Falls 86 (8.6) 23 (16.2) 66 (14.7) 59 (12.7) 2 (6.1) 12 (29.3) 17 (12.7) 265 (11.7)

Sport 129 (13.0) 7 (4.9) 41 (9.1) 55 (11.9) 5 (15.1) 4 (9.7) 13 (9.7) 254 (11.3)

Other 48 (4.8) 9 (6.4) 37 (8.3) 29 (6.3) 4 (12.1) 9 (22.0) 16 (11.9) 152 (6.7)

Unknown 9 (0.9) 2 (1.4) 4 (0.9) 2 (0.4) 0 (0.0) 0 (0.0) 1 (0.8) 18 (0.8)

Total for fracture typed 996 (44.1) 142 (6.3) 449 (19.9) 463 (20.5) 33 (1.5) 41 (1.8) 134 (5.9) 2,258 (100.0)

Treatment Conservative 245 (24.6) 51 (36.4) 212 (47.2) 186 (40.4) 9 (27.3) 31 (75.6) 68 (50.7) 802 (35.6)

Surgical fixation 736 (74.0) 82 (58.6) 223 (49.7) 218 (47.4) 22 (66.7) 5 (12.2) 54 (40.3) 1,340 (59.5)

Other 14 (1.4) 7 (5.0) 14 (3.1) 56 (12.2) 2 (6.1) 5 (12.2) 12 (9.0) 110 (4.9)

a6 patients had missing treatment data; totals do not sum to 1975 because some individuals experienced more than one type of injury; bInterpersonal violence; cRoad traffic accident;d Row percent

Table 5: Number of cases by reported primary cause, number of cases by fracture type, and number of cases by treatment type over the study period (percentages indicated in brackets unless otherwise indicated)

Primary Cause Period

2004–2006 2007–2009 2010–2013 IPV 296 (53.5) 293 (56.6) 487 (53.9)

RTA 96 (17.3) 69 (13.3) 121 (13.5)

Falls 49 (8.8) 58 (11.2) 123 (13.6)

Sport 73 (13.2) 65 (12.5) 97 (10.7)

Unknown 1 (0.2) 3 (0.6) 13 (1.4)

Other 39 (7.0) 30 (5.8) 62 (6.9)

Fracture Type Mandible 252 (39.4) 290 (49.1) 451 (43.9)

Maxilla 32 (5.0) 38 (6.4) 65 (6.3)

Orbit 120 (18.8) 98 (16.6) 230 (22.4)

Zygoma 166 (26.0) 111 (18.8) 196 (19.1)

Dentoalveolar 14 (2.2) 6 (1.0) 22 (2.1)

Other 55 (8.6) 48 (8.1) 64 (6.2)

Treatment Type Plating 293 (52.9) 344 (67.5) 507 (56.1)

Conservative 232 (41.9) 139 (27.3) 353 (39.1)

Other 29 (5.2) 27 (5.2) 43 (4.8) to 2013 for that catchment. This enabled person-years at risk to 45.5 per 100,000 rates to be calculated and compared to person-years at risk. Rates for IPV those reported by Buchanan et al (2005)1 increased over time from 20.7 per 100,000 by estimating the population at risk. We person-years at risk (95% CI = 18.5–23.2) used STATA (StataCorp LP, Texas, US) to to 24.5 per 100,000 person-years at risk calculate the rate ratios and their exact 95% (95% CI = 22.4–26.8). Facial fractures due confidence intervals. to RTA, sport and falls remained similar. Table 3 shows the percentage of primary Results reported cause by gender. Males accounted Table 1 lists fracture demographic char- for 77.2% of RTA-related fractures and acteristics and aetiological data. A total 87.2% of all fractures related to IPV and of 1,975 patients presented with maxil- sport. The highest prevalence for primary lofacial fractures. The male-to-female reported cause in females was falls (39.6%). ratio observed was approximately 4:1. A total of 2,258 fractures were recorded Those in the 20–29-year-old age group among the 1,975 individuals (Table 4). commonly presented with facial frac- Mandibular fractures were the most tures (38.4% of all cases). IPV was the common, with 996 individuals presenting. most common fracture aetiology (54.5%), Most of these were at a single site (54.5%). followed by RTA (14.5%), and sport The zygomatic complex was the next most (11.9%). Falls were the most common common fracture (463) followed by orbital cause in the oldest female age group (50+ fractures (449). The rate of orbital fractures years), while IPV was the most common rose from 8.4 to 11.6 per 100,000 person- among all other ages. years at risk. Other fracture types were less Table 2 shows that the rate of facial common. Most patients (87%) presented fractures increased from 38.8 per 100,000 with only one type of fracture; 8.8% had two

types of fracture (such as of the orbital floor the previous study at this unit.1 Although and zygoma); and 4.2% presented with three. the Department of Maxillofacial and Oral Table 5 presents data on how cases were Surgery at Waikato Hospital services treated over the study period. Over half a substantial and diverse population (58.2%) of all facial fractures were surgi- catchment, the findings cannot be gener- cally fixated and over one-third (36.8%) alised to all of New Zealand. Males aged were treated conservatively. The number of 20–39 most frequently presented with facial cases requiring surgical fixation increased fractures; this is a pattern which has been from 2004–2006 to 2007–2009 (p<0.001) but observed previously at this OMS unit and 1,2,4,5,7 no overall trend was observed. others around New Zealand. The rate of IPV-related maxillofacial trauma also Discussion increased at this tertiary centre, up from This study was a retrospective clinical 12.7 (CI 11.9–13.5) per 100,000 person-years audit of patients who presented, over a at risk in 1989–2000 to 24.5 (CI 22.4–26.8) by 10-year period, with facial fractures to a 2004–2013. Currently, it is almost twice the tertiary trauma hospital Oral and Maxillo- rate of the 1989–2000 period (p<0.0001) and facial Surgery (OMS) unit. Fractures were this is of significant concern.1 most common in young men. The decrease in the number of RTA-re- We observed an increase in the rate of lated fractures observed is in keeping with IPV-related fractures, orbital fractures, observations from other OMS units around fractures due to falls, and the total rate of New Zealand.1,2,5,6 Interestingly, despite this, facial fractures. The rate of RTA-related we did not observe an overall decrease in cases fell slightly, while sports-related high-velocity injury fracture patterns, such fractures were similar. as Le Fort fractures, because there was an Missing age, sex, and treatment data were increase in Le Fort 1 fractures caused by noted as part of imperfect data entry into IPV. Although facial fracture rates due to the database. Fortunately, these were low falls were similar to the previous study, they in number and unlikely to have affected continued to be common in the female 50+ 1,20 the findings. Having missing data remains age group. It is clear that falls continue a risk with such analyses of routinely-col- to be a significant cause of maxillofacial lected data: treating clinicians have many trauma in older adults, and this is likely to responsibilities and distractions, and grow as New Zealand’s older-adult popu- these are likely to compromise data entry lation increases. on occasion. Another limitation of this Unexpectedly, the rate of orbital fractures study was that information on ethnicity, observed was 5.1 times higher (p<0.0001) or alcohol consumption prior to injury, than that seen in the previous study at this was not available. The original purpose unit.1 This was due to a large rise in orbital of the database was to provide hospital fractures caused by IPV, RTA, and falls since management with an indication of the the 1989–2000 period.1 Similar proportions departmental workload. Despite this, of orbital fracture cases have been reported sporadic data were collected on ethnicity in other New Zealand studies.3,4 The and alcohol, but this information was not increase in the orbital fracture rate is likely of sufficient quality to be included in this due to increased detection and referral analysis. However, many previous studies from secondary centres. have reported alcohol consumption to be a Initially, the proportion of cases conser- strong contributing factor to facial frac- vatively managed was similar to figures tures caused by IPV and RTA, especially reported by other units, but it decreased 1,4,5,7,8 in the 15–39 age group. Previously, towards the end of the study period.4,5,7,8 Māori have been found to be over-repre- The proportion of cases requiring surgery sented in facial fracture presentations at in New Zealand does vary due to differ- 1 this particular tertiary hospital. There is no ences in the local populations and in the reason to suspect that any change in pattern available service-mix. The proportion of has occurred since the earlier study. fractures requiring internal fixation did Important trends were identified and can increase from previously (about 60%, up be directly compared with findings from from 49%) and is higher than that seen in

another New Zealand hospital during a and drug use, but also the main culprits, similar period.1,4 This probably has a multi- namely young males. factorial origin, reflecting an increase in Further studies are needed to investigate proportion of displaced fractures, changes the role of alcohol and drugs (especially in surgical staff, and a rise in serious inter- methamphetamine) in this population personal violence in the Waikato region. presenting with facial fractures. Facial fractures due to interpersonal violence have continued to rise, while Conclusion those caused by road traffic accidents have The rate of IPV related maxillofacial continued to decline. However, the ever-in- fractures presenting to this tertiary centre creasing number of facial fractures places has continued to increase throughout the pressure on staffing levels, ward beds, study period and is now at almost double theatre availability, and hospital funding. the rate since the 1998-2000 period. It is apparent that the level of interper- It continues to be the dominant cause sonal violence has worsened since the of injury, while RTA-related fractures previous study, and remains an important are decreasing. The ever-increasing public health issue in the Waikato region. rate of facial fractures presenting to Programmes such as the Violence Inter- Waikato Hospital places significant vention Programme (VIP) and ‘It’s not demands on scarce clinical resources, OK’, which have been running across such as operating theatre time and New Zealand during the study period, staffing numbers. An increase in theatre appear to have had no impact on reducing access for maxillofacial doctors and IPV-related maxillofacial fractures in the additional staff in oral and maxillofacial Waikato region.21 With such a clear impact surgery departments is essential to deal on current and future hospital resources adequately with this problem. IPV is an and staffing, public health intervention escalating cause of facial fractures that is required. This should target contrib- requires urgent and interventional public utory factors, such as alcohol consumption health prevention strategies.

Competing interests: Nil Acknowledgements: The authors thank Mr Steve Evans, Mr Brian Whitley and Mr Simon Lou, Consultant Oral and Maxillofacial Surgeons, Waikato Hospital, who contributed to the surgical treatment of the patients, and Dr Dalice Sim, University of Otago, Wellington School of Medicine, for assisting with the statistical analysis of data. Author information: Blake Moore, 5th-year medical student, University of Otago, Wellington School of Medicine, Wellington, New Zealand; Ryan Smit, 5th year medical student, University of Otago, Wellington School of Medicine, Wellington, New Zealand; Angus Colquhoun, Consultant Oral and Maxillofacial Surgeon, Department of Maxillofacial and Oral Surgery, Waikato Hospital, New Zealand; W. Murray Thomson, Professor of Dental Epidemiology and Public Health, Department of Oral Sciences, University of Otago Faculty of Dentistry, Dunedin, New Zealand. Corresponding author: Blake Moore, 5th year medical student, University of Otago, Wellington School of Medicine, Wellington, New Zealand [email protected] URL: www.nzma.org.nz/journal/read-the-journal/all-issues/2010-2019/2015/vol-128-no-1426-4- december-2015/6752

REFERENCES: 1. Buchanan J, Colquhoun fractures. N Z Med J. 5-Year Study of 237 A, Friedlander L, Evans 2008; 121(1271):15-23. Patients. J Oral Maxillofac S, Whitley B, Thomson 9. Timoney N, Saiveau M, Surg. 2003;61:61-64. WM. Maxillofacial Pinsolle J, Shepherd J. 16. Tanaka N, Tomitsuka K, fractures at Waikato A comparative study of Shionoya K, et al. Aetiology Hospital, New Zealand: maxillo-facial trauma in of maxillofacial fracture. 1989 to 2000. N Z Med J. Bristol and Bordeaux. J Br J Oral Maxillofac 2005;118(1217):U1529. Craniomaxillofac Surg. Surg. 1994; 32:19–23. 2. Kieser J, Stephenson 1990;18:154–157. 17. Zachariades N, Papavas- S, Liston PN, Tong DC, 10. Kontio R, Suuronen R, siliou D. The pattern and Langley JD. Serious facial Ponkkonen H et al. Have aetiology of maxillofacial fractures in New Zealand the causes of maxillofacial injuries in Greece. A from 1979 to 1998. Int. fractures changed over the retrospective study of 25 J. Oral Maxillofac. Surg. last 16 years in Finland? years and a comparison 2002;31:206–209. An epidemiological study with other countries. J 3. Love R and Ponnambalam of 725 fractures. Dent Craniomaxillofac Surg. Y. Dental and maxillofacial Traumatol. 2005;21:14–19. 1990;18:251–254. skeletal injuries seen at 11. van Beek GJ and Merkx 18. van den Bergha B, the University of Otago CA. Changes in the Karagozoglua K, School of Dentistry, pattern of fractures of the Heymans M, Forouzanfar New Zealand 2000-2004. maxillofacial skeleton. T. Aetiology and inci- Dental Traumatology. Int J Oral Maxillofac dence of maxillofacial 2008;24:170–176. Surg. 1999;28:424–428. trauma in Amsterdam: 4. Lee K. Interpersonal 12. Brasileiro B and Passeri A retrospective anal- Violence and Facial Frac- L. Epidemiological ysis of 579 patients. tures. J Oral Maxillofac analysis of maxillofacial Journal of Cranio-Max- Surg. 2009;67:1878-1883. fractures in Brazil: A illo-Facial Surgery. 5. Lee K, Snape L, Steen- 5-year prospective study. 2012;40:e165–e169. berg L, Worthington J. Oral Surg Oral Med 19. Hagan EH, Huelke DE. Comparison between Oral Pathol Oral Radiol An analysis of 319 case interpersonal violence and Endod. 2006;102:28-34. reports of mandibular motor vehicle accidents in 13. Chrcanovic BR, Freire- fractures. J Oral Surg. the aetiology of maxillo- Maia B, Souza LN et al. 1961;19:93-104. facial fractures. ANZ J. Facial fractures: a 1-year 20. Thomson WM, Stephenson Surg. 2007;77:695–698. retrospective study S, Kieser J, Langley JD. 6. Koorey AJ, Marshall SW, in a hospital in Belo Dental and maxillofacial Treasure ET, Langley Horizonte. Braz Oral injuries among older New JD. Incidence of facial Res. 2004;18: 322–328. Zealanders during the fractures resulting in 14. Weihsin H, Thadani S, 1990s. Int. J. Oral Maxillo- hospitalisation in New Agrawal M, Tailor S. fac. Surg. 2003;32:201–205. Zealand from 1979 to Causes and incidence of 21. Koziol-McLain J, McLean 1988. Int. J. Oral Maxillo- maxillofacial injuries in C, Garrett N. Hospital fac. Surg. 1992;21:77-79. India: 12-year retrospec- Responsiveness to Family 7. Lee K and Antoun J. tive study of 4437 patients Violence: 108 Month Zygomatic fractures in a tertiary hospital in Follow-Up Evaluation. presenting to a tertiary Gujarat. British Journal Ministry of Health VIP trauma centre, 1996-2006. of Oral and Maxillofacial report 2013. http://www. New Zealand Dental Surgery. 2014;52:693–696. aut.ac.nz/__data/assets/ Journal. 2009;105(1):4-7. 15. Motamedi M. An pdf_file/0003/447285/ 8. Lee K and Snape L. Role Assessment of Maxil- WEB_108M-VIP-FU- of alcohol in maxillofacial lofacial Fractures:A REPORT-2013.pdf

A bug in the ointment: topical antimicrobial usage and resistance in New Zealand Deborah A Williamson, Stephen R Ritchie, Emma Best, Arlo Upton, Alison Leversha, Alesha Smith, Mark G Thomas

ABSTRACT New Zealand has unenviably high rates of bacterial resistance to topical antimicrobials. In this Viewpoint, we review the history and usage of topical antimicrobials in New Zealand, and suggest some strategies to mitigate further increases in antimicrobial resistance to topical agents.

lence of high-level mupirocin resistance in S. Background aureus from 14.2% in 2000, to 8.3% in 2014.5 Antimicrobial resistance has been Interestingly, the authors of a 2003 study described as a “crisis for the health and describing mupirocin resistance in New wealth of nations”.1 One of the key strat- Zealand concluded that: egies to mitigate this public health crisis is to “In cautioning against the use of ensure that existing antimicrobials are used mupirocin, we do not advocate using responsibly and judiciously. In general, New fusidic acid topically as an alter- Zealand has relatively low rates of antimi- native. Resistance to this topical crobial resistance threats deemed as ‘critical’ agent is reported, and unlike mupi- or ‘urgent’ by the Centers for Disease Control rocin, it is available in oral and and Prevention.2 However, this may reflect intravenous formulations that are relatively low levels of consumption of used for treatment of multiresistant many antimicrobials in previous decades, S. aureus infections”.4 rather than our relatively high levels of Similarly, an Australian commentary in antimicrobial consumption in more recent 2006 on fusidic acid use stated: years.3 In particular, high usage of topical antimicrobial agents over the past three “Ensuring that… the use of topical decades in New Zealand has resulted in an fusidic acid is either abolished or ill-fated series of national population-level restricted will be vital if we are to experiments, which clearly illustrate prevent the loss of this potentially the relationships between antimicrobial useful agent”, and “Common sense consumption and resistance. Throughout would suggest that antibiotics used the 1990s, the topical antimicrobial agent topically should be ones that are not 6 mupirocin (Bactroban©) was available to used systemically”. purchase ‘over-the-counter’ (OTC), which led Despite these unambiguous warnings, to very high levels of use, and subsequent regulatory changes and the promulgation high rates of resistance, such that by 2000, of guidelines promoting the use of topical approximately 14% of S. aureus isolates fusidic acid ointment and cream contributed displayed high-level resistance to mupi- to a significant increase in topical fusidic rocin.4 From April, 2000, regulatory changes acid dispensing in New Zealand throughout meant that mupirocin could be obtained the 2000s (Figure 1), with an associated by ‘prescription only’, with a subsequent increase in the prevalence of resistance in S. decrease in usage, and a fall in the preva- aureus from 17% in 1999, to 28% in 2013.7 At

Figure 1: Community dispensing rates per 1,000 population for topical fusidic acid and mupirocin in the New Zealand community setting, January, 1993,–August, 2013. Reproduced from Reference 7.

present, topical fusidic acid is available by widespread use of topical fusidic acid.13-15 prescription only, although it is fully subsi- In general, Streptococcus pyogenes, the dised by the New Zealand Ministry of Health, other pathogen commonly associated with unlike mupirocin, which is only partially impetigo, is less susceptible to fusidic acid subsidised. than S. aureus.16 The evidence for and against One of the largest randomised control topical antimicrobial use trials (RCTs) assessing the efficacy of topical fusidic acid vs placebo in the treatment of Theoretically, the use of topical anti- mild impetigo, conducted in the Nether- microbials is an attractive option to treat lands between 1999 and 2000, found that minor skin ailments. Topical application cure rates after one week of treatment allows delivery of high concentrations of with topical fusidic acid were significantly antimicrobial at the site of infection, while higher than with placebo (55% vs 13%, odds minimising systemic absorption. In practice ratio [OR] 12.6, 95% confidence interval however, topical antibiotic use has long [CI], 5.0–31.5).17 However, this difference been recognised as a very efficient method reduced over time, with 92% of treated of rapidly promoting the emergence patients displaying cure at 28 days, vs 88% and proliferation of antibiotic resistant of patients in the placebo arm. It also must microbes.6 Furthermore, evidence-based be noted that no fusidic acid resistance prescribing supports the use of topical was detected in S. aureus isolates from antimicrobial agents for only a few specific this study population, meaning that these indications, including nasal eradication of S. study findings are not directly applicable aureus, treatment of acne, and treatment of to the New Zealand setting, where contem- 8-10 mild impetigo. porary fusidic acid resistance rates are Despite concerns about efficacy and high. In addition, another RCT conducted the promotion of even higher rates of in the UK, Germany and Sweden in 1994, resistance, fusidic acid remains the recom- found no statistically significant difference mended agent in New Zealand for the in cure rates between topical fusidic acid empiric treatment of impetigo.11,12 Impor- and hydrogen peroxide in the treatment tantly, rates of resistance to fusidic acid of localised impetigo (82% vs 72%, respec- in S. aureus remain comparatively lower tively).18 Again, caution should be exercised in countries that have not adopted the when extrapolating these results to the New

Figure 2: Community dispensing rates per 1,000 population for topical fusidic acid in the New Zealand community setting stratified by age group, January, 2006,–August, 2013.

Figure 3: Community dispensing rates per 1,000 population for topical fusidic acid in the New Zealand community setting stratified by ethnicity, January, 2006,–August, 2013.

Figure 4: Community dispensing rates per 1,000 population for topical fusidic acid in the New Zealand community setting stratified by geographic region, January, 2006–August, 2013. (Northern = Northland DHB, Waitemata DHB, Auckland DHB, Counties Manukau DHB; MidCentral = Waikato DHB, Lakes DHB, Bay of Plenty DHB, Tairawhiti DHB, Taranaki DHB; Central = Hawkes Bay DHB, MidCentral DHB, Whan- ganui DHB, Capital and Coast DHB, Hutt DHB, Wairarapa DHB; Southern = Nelson Marlborough DHB, West Coast DHB, Canterbury DHB, South Canterbury DHB, Southern DHB).

Zealand setting, as rates of fusidic acid resis- with topical antimicrobial usage.19 These tance in these countries are markedly lower authors hypothesised that in such instances, than New Zealand.14 However, despite the topical antimicrobials were being used limited application of overseas findings to as post-operative wound ‘prophylaxis’ the New Zealand setting, such studies have following minor surgery, a practice that is been used as the basis for guidelines that not supported by available evidence.20,21 actively recommend topical fusidic acid Information on the demographics and 11,12 in the empiric treatment of impetigo. geographic variation of antimicrobial Importantly, there are no published studies usage in a population is essential in under- comparing the use of topical fusidic acid standing how and why antimicrobials are vs placebo, or vs antiseptic treatment for utilised, and identifying potential areas impetigo in high prevalence resistance for reduction in usage. Information on all settings, such as New Zealand. In particular, community prescriptions in New Zealand it is not yet known whether topical are maintained in a central data warehouse, hydrogen peroxide is a feasible alternative the ‘Pharmaceutical Collection’. Data from for New Zealand children compared to this collection between January, 2006, and topical fusidic acid. August, 2013, demonstrates that the highest Demographics of topical fusidic rates of topical fusidic acid dispensing were acid use in New Zealand in the under-5 year age group, followed by the 75 year and over age group (Figure 2). In addition to therapeutic usage, data suggest that, in some settings, topical When stratified by ethnicity, the highest antimicrobials may also be used prophy- rates of dispensing were in Māori and lactically, particularly in elderly patients. Pacific Peoples (Figure 3), and when strat- For example, a study from the US assessing ified by geographic region, the highest rates national usage of topical antimicrobials of dispensing were in the Northern region found that 40% of all topical antimicrobial of New Zealand (Figure 4). usage was in the over-50 age group, with These dispensing patterns are consistent benign or malignant skin neoplasms being with recent work showing the high rates the most common diagnosis associated of skin infections in Pacific and Māori

children,22 and further emphasise the prescribing what may be regarded as a considerable burden of skin disease in benign treatment. these groups. In addition, the high rates of Collective action requires dispensing in the Northern region reflect the higher incidence of skin disease in this collective responsibility It is clear from available data that the rate region, which has the highest population of of fusidic acid resistance in New Zealand Māori and Pacific Peoples in New Zealand. Furthermore, the relatively high rates of is one of the highest in the developed topical fusidic acid usage in the over-75 world, and that high levels of usage have year age group are concerning, partic- contributed to proliferation of the AK3 ularly given the limited evidence-based MRSA clone. It is also important to note that indications for prescribing topical anti- a considerable proportion of topical antimicrobials in older age groups.19 To date microbial usage in New Zealand may be however, there are no available data on considered ‘appropriate’, particularly given the clinical indications for topical antimi- the high burden of childhood skin infec- 22,24 crobial prescribing in elderly patients in tions in our setting. However, in the face New Zealand. Such information is critical of high bacterial resistance, we question for determining whether current usage of the value of continuing to recommend topical antimicrobials is clinically indicated, topical fusidic acid as empiric therapy in and identifying strategies to reduce inap- New Zealand, and suggest a multipronged propriate prescribing. approach aimed specifically at reducing rates of resistance: Collateral damage caused by 1. Consistent, evidence-based, national high levels of fusidic acid usage guidelines around the appropriate in New Zealand use of topical antimicrobials. Recent data suggest that, as might be 2. Reduce the volume of agent expected, the high usage of topical fusidic dispensed to patients (eg, a 5g tube acid in New Zealand is driving the increase instead of a 15g tube). in fusidic acid resistant S. aureus clones.5,7,23 Of specific concern is the emergence of a 3. Regulatory measures around the use fusidic acid-resistant community-associated of topical fusidic acid, such as moving methicillin-resistant S. aureus (MRSA) clone, to ‘specialist-only’ prescribing in the known in New Zealand as the ‘AK3’ clone.23 elderly. This clone has rapidly become the most 4. Improved education to primary care common type of MRSA causing illness in practitioners about evidence-based New Zealand.5 Genomic data indicates that prescribing of topical antimicrobials, the gene conferring fusidic acid resistance particularly in elderly patients. (fusC) and the gene conferring methicillin 5. Clear messaging to the public about resistance (mecA) are located together the importance of not sharing topical on the same mobile genetic element.7 In antimicrobials amongst a household, simple terms, this means that large-scale and discarding any remaining topical use of topical fusidic acid has favoured the agent once the treatment course has proliferation of the AK3 MRSA clone, and been completed. has provided a ‘helping hand’ in allowing this clone to become established in New 6. Robust clinical trials, conducted in Zealand. In addition, a recent national study a setting with a high prevalence of of antimicrobial resistance in New Zealand resistance to topical agents, assessing found that 36% of all fusidic acid-resistant the clinical utility of antiseptic agents methicillin-susceptible S. aureus (MSSA) in the treatment of localised impetigo. strains were also resistant to mupirocin, A key first step would be identification highlighting the potential for treatment and gathering of relevant stakeholders, with one antimicrobial to select for multi- and formation of a clear ‘road-map’ to resistant bacterial clones.5 In this context, address this significant problem. These it is important for practitioners to be aware stakeholders should include prescribers, the of the wider ecological implications (or Ministry of Health, PHARMAC and patient ‘collateral damage’) that can occur when representatives. New Zealand has already

had considerable success in reducing similar concerted approach, involving rates of topical antimicrobial resistance prescribers, policy makers, and patients, is encountered in S. aureus isolates. This is urgently required to tackle our unenviably highlighted by the reversal in mupirocin high rates of fusidic acid resistance, and resistance in New Zealand over the past confront our over prescription of topical 15 years, which was, in part, due to both antimicrobial agents. educational and regulatory measures. A

Competing interests: Nil Author information: Deborah A Williamson, Institute of Environmental Science and Research, Wellington, New Zealand, and University of Otago, Wellington, New Zealand; Stephen R Ritchie, Faculty of Medical and Health Sciences, University of Auckland, Auckland, New Zealand; Emma Best, Faculty of Medical and Health Sciences, University of Auckland, Auckland, New Zealand; Arlo Upton, LabTests, Mount Wellington, Auckland, New Zealand; Alison Leversha, Faculty of Medical and Health Sciences, University of Auckland, Auckland, New Zealand; Alesha Smith, School of Pharmacy, University of Otago, New Zealand and bpacnz, Dunedin, New Zealand; Mark G Thomas, Faculty of Medical and Health Sciences, University of Auckland, Auckland, New Zealand. Corresponding author: Dr Deborah Williamson, Department of Pathology and Molecular Medicine, PO Box 7343, Wellington South, 6242, New Zealand [email protected] URL: www.nzma.org.nz/journal/read-the-journal/all-issues/2010-2019/2015/vol-128-no-1426-4- december-2015/6753

REFERENCES: 1. ‘Review on Antimicrobial nan H. Mupirocin and Dis. 2006; 42(3):394-400. Resistance. Antimicrobial Staphylococcus aureus: A 7. Williamson DA, Monecke Resistance: Tackling recent paradigm of emerg- S, Heffernan H et al. A a Crisis for the Health ing antibiotic resistance. cautionary tale: High and Wealth of Nations. J Antimicrob Chemother. usage of topical fusidic 2014.’ Available at: www. 2003; 51: 613-7. acid and rapid clonal amr-review.org, last 5. Heffernan H, Bakker expansion of fusidic accessed 25th April, 2015 S, Woodhouse R, Dyet acid-resistant Staphylo- 2. Centers for Disease K, Williamson DA. coccus aureus. Clin Infect Control and Preven- Demographics, antimi- Dis. 2014; 59(10):1451-4. tion (CDC). Antibiotic crobial susceptibility and 8. Verhoeven PO, Gagnaire resistance threats in the molecular epidemiology J, Botelho-Nevers E United States. 2013. of Staphylococcus aureus et al. Detection and Available at: http://www. in New Zealand, 2014 clinical relevance of cdc.gov/drugresistance/ (https://surv.esr.cri. Staphylococcus aureus pdf/ar-threats-2013-508. nz/PDF_surveillance/ nasal carriage: An update. pdf, last accessed Antimicrobial/Staph Expert Rev Anti Infect 25th April, 2015 /2104Saureussurveyre- Ther. 2014; 12: 75-89. 3. Thomas MG, Smith AJ, port.pdf, last accessed 9. Koning S, van der Sande Tilyard M. Rising anti- 20th April, 2015) R, Verhagen AP et al. microbial resistance: a 6. Howden BP, Grayson ML. Interventions for impetigo. strong reason to reduce Dumb and dumber--the Cochrane Database Syst excessive antimicrobial potential waste of a useful Rev. 2012; 1: CD003261. consumption in New antistaphylococcal agent: 10. Thornton Spann C, Taylor Zealand. N Z Med J. Emerging fusidic acid SC, Weinberg JM. Topical 2014;127(1394):72-84 resistance in Staphylo- antimicrobial agents 4. Upton A, Lang S, Heffer- coccus aureus. Clin Infect in dermatology. Clin

Dermatol. 2003; 21: 70-7. mechanisms among indications. Dermatol 11. Vogel A, Lennon D, Gray Staphylococcus spp. Surg. 2011; 37: 1427-33. S et al. Registered nurse isolated in North America 20. McHugh SM, Collins CJ, assessment and treatment and Australia, 2007-2008. Corrigan MA, Hill AD, of skin sepsis in New Antimicrob Agents Chemo- Humphreys H. The role Zealand schools: The ther. 2010; 54(9):3614-7. of topical antibiotics used development of protocols. 16. Jones RN, Mendes RE, as prophylaxis in surgical N Z Med J. 2013; 126: 27-38. Sader HS et al. In vitro site infection prevention. antimicrobial findings for 12. Best Practice Advisory J Antimicrob Chemother. fusidic acid tested against Committee. Management 2011 Apr;66(4):693-701. contemporary (2008-2009) of impetigo. Available at 21. White R, Cooper R, Kings- gram-positive organisms http://Www.bpac.org.nz/ ley A. Wound colonization collected in the United BPJ/2009/february/docs/ and infection: the role of States. Clin Infect Dis. bpj19_impetigo_pages_8- topical antimicrobials. Br 2011; 52 Suppl 7: S477-86. 11.pdf. 2009, last accessed J Nurs 2001; 10(9): 563-78 17. Koning S, van Suijle- 15th June, 2015 22. Williamson DA, Zhang kom-Smit LW, Nouwen 13. Coombs GW1, Daly J, Ritchie SR et al. JL, Verduin CM, Bernsen DA2, Pearson JC1, et Staphylococcus aureus RM, Oranje AP, Thomas al. Community-onset infections in New Zealand, S, van der Wouden JC. Staphylococcus aureus 2000-2011. Emerg Infect Fusidic acid cream in the Surveillance Programme treatment of impetigo in Dis. 2014; 20: 1156-61. annual report, 2012. general practice: double 23. Williamson DA, Roberts Commun Dis Intell Q blind randomised placebo SA, Ritchie SR et al. Clinical Rep. 2014; 38(1):E59-69. controlled trial. BMJ. and molecular epidemiol- 14. Castanheira M, Watters 2002; 324(7331):203-6 ogy of methicillin-resistant AA, Mendes RE et al. 18. Christensen OB, Anehus Staphylococcus aureus Occurrence and molec- S. Hydrogen peroxide in New Zealand: Rapid ular characterization of cream: an alternative emergence of sequence fusidic acid resistance to topical antibiotics in type 5 (ST5)-SCCmec-IV as mechanisms among the treatment of impe- the dominant communi- Staphylococcus spp. tigo contagiosa. Acta ty-associated MRSA clone. from European countries Derm Venereol. PLoS One. 2013; 8: e62020. (2008). J Antimicrob 1994; 74(6): 460-2 24. O’Sullivan CE, Baker Chemother. 2010; 65: 1353- 19. Lapolla WJ, Levender MG, Zhang J. Increasing 15. Castanheira M, Watters MM, Davis SA et al. hospitalizations for AA, Bell JM, Turnidge Topical antibiotic trends serious skin infections in JD, Jones RN. Fusidic from 1993 to 2007: Use New Zealand children, acid resistance rates and of topical antibiotics 1990-2007. Epidemiol prevalence of resistance for non-evidence-based Infect. 2011; 139: 1794-804.

Successful conservative management of campylobacter cholecystitis occurring post chemotherapy and rituximab: a rare disease entity Ajay Gupta, Louise Teo

ABSTRACT Campylobacter jejuni is commonly associated with gastroenteritis, but extremely few reports worldwide link it acute cholecystitis. These infectious complications can assume menacing proportions in the immunocompromised and need careful management. We present a report of such a case from Australia, successfully managed conservatively, without surgery.

ampylobacter jejuni is commonly to the onset of symptoms. On examination, associated with gastroenteritis,1-3 he was dehydrated. He had a temperature but extremely few reports link it of 39°C, associated with mild chills, and C 4-8 with acute cholecystitis. These infectious right upper-quadrant tenderness with complications can assume menacing positive Murphy’s sign. proportions in the immunocompromised, He was initially started on piperacil- and need careful management. We present lin-tazobactam 4.5g and metronidazole a report of such a case from Australia, 500mg intravenously every 8 hours for successfully managed conservatively, presumed febrile neutropenia. The patient without surgery. had received peg-filgrastim, post chemotherapy and had an elevated neutrophil Case count of 14.72x109/L. Abdominal ultrasound A 65-year-old gentleman having non-Hod- reported features of acute cholecystitis, gkin’s lymphoma Stage 4 was started on with gall bladder wall thickening and Rituximab, cyclophosphamide, vincristine enlargement, mobile gallstones and small and prednisolone (RCHOP). On Day 11 amounts of pericholecystic fluid. The region of his second cycle of chemotherapy, he was tender to probe pressure (sonographic presented to the emergency department Murphy’s sign positive). Hepatic steatosis with a three-day history of severe right was also present. The common bile duct upper-abdominal pain, vomiting, high- was not dilated, nor the biliary tree. There grade fever and diarrhea. The diarrhea was no biliary sludge. The stool microscopic was florid—16–18 episodes in a day, occa- examination revealed pus cells and red sionally bloody, but mostly watery and blood cells, suggesting infectious colitis. associated with tenesmus. He had cooked A surgical consult was taken, but owing chicken stew approximately 36 hours prior to his weak general condition, the patient

was deemed unfit for any surgical inter- We do believe that successful conser- vention by the surgeons and was managed vative management is possible in patients conservatively. deemed unfit for surgery owing to poor He was managed conservatively with general condition. Another alternative is intravenous fluids and small amounts of percutaneous cholecystostomy. clear fluid orally for the first 72 hours, Percutaneous cholecystostomy is usually followed by institution of a low-fat, semi- indicated in patients who fail an initial solid diet and oral rehydration therapy. trial of antibiotic therapy. However, On the fifth day, the stool culture was gallbladder drainage by percutaneous reported positive for Campylobacter jejuni. cholecystostomy in conjunction with anti- Parenteral antibiotics were stopped and biotics may be used as initial treatment for oral ciprofloxacin 750mg twice daily very ill patients (ie, intensive care unit). was started. Following this, his condition The procedure is not without risk and in improved, and his clinical symptoms one retrospective review that included resolved over the next one week. His right 1,918 patients, 30-day mortality after upper-quadrant pain also subsided, coin- percutaneous cholecystostomy was 15.4 ciding with the resolution of his diarrhea. %, but only 4.5 % for cholecystectomy,9 The patient was discharged home on the difference being likely due to patient ciprofloxacin for a total of two weeks of oral selection bias. Minor complications of therapy. An ultrasound, repeated 10 days percutaneous drainage include bleeding, later, demonstrated complete resolution of catheter blockage and dislodgement the acute cholecystitis. (10–15%), and failure to resolve the acute cholecystitis (10%). 10 However, our patient Campylobacters are common commensals responded promptly to change of therapy in the gastrointestinal tract of animals, to ciprofloxacin, thus obviating the need especially poultry. Campylobacter jejuni is known to cause gastroenteritis, colitis, for such a procedure. septicemia, peritonitis, pancreatitis and Mobile gall stones were present in our gastrointestinal hemorrhage besides many case, but the liver function tests were extra-intestinal complications.1-3 However, normal. There was no biliary sludge or bile association with acute cholecystitis is duct dilatation on ultrasound. There was exceedingly rare.4-8 an extremely strong temporal relationship There have been 16 reported cases4-8 between both the onset and the resolution (15 published and 1 unpublished poster of cholecystitis and gastroenteritis. As report6) worldwide (none from Australia/ mentioned above, there was rapid and Oceania) that have linked Campylobacter simultaneous resolution of cholecystitis, infection with cholecystitis. The clinical along with the gastroenteritis upon insti- presentation included abdominal pain tution of C.jejuni specific antimicrobial in all, fever in nine, vomiting in eight, therapy with ciprofloxacin. Diarrhea and jaundice in two, septic shock with hypo- pre-existing gall stones, as seen in our case, tension in two and mortality in one case have been described in Campylobacter with advanced hepatocellular carcinoma. cholecystitis, as detailed in the review Diarrhea and pre-existing gall stones, as of literature above. Also, at least in one seen in our case, were described in six reported case, the diagnosis of Campylo- and seven cases, respectively.4-8 bacter infection was based on initial stool Three of the cases were treated conserva- culture rather than bile cultures post chole- 8 tively only with antibiotics, while the rest cystectomy. All these facts, taken together, were managed with cholecystectomy. For strongly argue for both conditions having Campylobacter infections, ciprofloxacin or a common etiology in the form of Campylo- macrolides are antibiotics of choice. Most bacter jejuni. notably, in one of the cases, the clinical In patients with lymphoma, use of Ritux- condition of the patient continued to dete- imab—a monoclonal antibody against CD riorate after cholecystectomy and resolved 20 positive B cells, adds to the immunosup- only upon institution of specific targeted pression and may have contributed to the antibiotic treatment against C.jejuni.7 severe manifestations seen in our case.

Studies suggest strong association coverage presumptively in such situations, between fatal outcome and prescription rather than wait for confirmatory investi- of a third-generation cephalosporin for gations so as to avoid serious complications patients with Campylobacter (except C.fetus) especially in the immunocompromised, in bacteremia, especially in the immunocom- whom there is greater likelihood of being promised.2,3 Thus, it would be prudent to deemed unfit for surgical interventions. provide anti-Campylobacter anti-microbial

Competing interests: Nil Author information: Ajay Gupta, Department of Medical Oncology, Asian Cancer Center and Ex Consultant, Medi- cal Oncology, Hervey Bay, Queensland, Australia; Louise Teo, Medical Oncology, Hervey Bay Hospital, Queensland, Australia. Corresponding author: Ajay Gupta, Department of Medical Oncology, Asian Cancer Center and Ex Consultant, Medical Oncology, Hervey Bay, Queensland, Australia [email protected] URL: www.nzma.org.nz/journal/read-the-journal/all-issues/2010-2019/2015/vol-128-no-1426-4- december-2015/6754

REFERENCES: 1. Kapperud G, Lassen J, cholecystitis with lithiasis. entity. BMJ Case Rep. Ostroff SM, Aasen S. Clin- Case report and literature 2010;2010:bcr1020092365. ical features of sporadic review. Clin Microbiol doi: 10.1136/ Campylobacter infections Infect 2003; 9: 970–2 bcr.10.2009.2365. in Norway. Scand J Infect 5. Takatsu M, Ichiyama S, 8. Udayakumar D, Sanaul- Dis 1992; 24:741. Nada T, et al. Campy- lah M. Campylobacter 2. Skirrow MB, Blaser lobacter fetus subsp. fetus cholecystitis. Int J Med MJ. Clinical aspects of cholecystitis in a patient Sci. 2009 :6:374-5. Campylobacter infection. with advanced hepatocel- 9. Abi-Haidar Y, Sanchez V, In: Campylobacter, 2nd lular carcinoma. Scand J Williams SA, Itani KM. ed, Nachamkin I, Blaser Infect Dis 1997; 29: 197–8. Revisiting percutaneous MJ (Eds), ASM Press, 6. Verzotti G, Muradbegovic cholecystostomy for acute Washington DC 2000. p.69. M, Schneider R. Acalculous cholecystitis based on a 3. Pacanowski J1, Lalande V, cholecystitis due to Campy- 10-year experience. Arch Lacombe K, etal. Campy- lobacter jejuni: should Surg 2012; 147:416-422. lobacter bacteremia: we operate? http://www. 10. Byrne MF, Suhocki P, clinical features and chirurgiekongress-poster. Mitchell RM, et al. Percu- factors associated with ch/fileadmin/files/docu- taneous cholecystostomy fatal outcome. Clin Infect ments/pdfs-2015/3759.pdf in patients with acute Dis. 2008 ;47:790-6. 7. Vaughan-Shaw PG, Rees cholecystitis: experience 4. Dakdouki GK, Araj GF, JR, White D. Campylobac- of 45 patients at a US Hussein M. Campylobacter ter jejuni cholecystitis: a referral center. J Am Coll jejuni: unusual cause of rare but significant clinical Surg 2003; 197:206-11.

Over five years on, we still don’t need a pilot bowel cancer screening programme: Please just get on with it! Guy Hingston

ear Sir, here? And why are we debating the cost of It is now over five years since expensive pharmaceuticals for metastatic 2 DI wrote to you stating that, “The bowel cancer? We’ve known for five years only pilot we need is someone with the that we can halve the number of left-sided courage and funding to roll out a national metastatic bowel cancer deaths by flexible … programme to save over 500 Kiwi sigmoidoscopic screening, which in turn lives each year!”1 The then quoted Level would halve the pharmaceutical bills! 1A prospective randomised evidence It’s a sad fact that New Zealand has confirming a 50% reduction in left-sided the worst bowel cancer statistics of any bowel cancer deaths from flexible sigmoid- ‘developed’ country, and it’s obviously oscopy screening2 has been ignored by most because we haven’t ‘developed’ a bowel in New Zealand, even though similar results cancer screening program. We need to have now come out of Norway.3 end this discrimination by region. Should I have just completed a short contract all New Zealanders between the age of for two DHBs to help take 200 people off 50–75 now shift to live in the Waitemata their colonoscopy waiting list, and I write area, so they are not further discriminated with concern, noting that only one of against by where they live? And why is a these people had decided to purchase a positive FOBT the most common reason for Faecal Occult Blood Test (FOBT) kit from a a colonoscopy in Australia, and the least pharmacy. She returned a positive test and common reason in New Zealand? we subsequently removed a large prema- I am hoping to turn 50 next year, and lignant polyp, thus probably preventing her am feeling guilty about planning to have succumbing prematurely to bowel cancer. my first screening colonoscopy, noting that Is it acceptable that the Ministry of Health many of my colleagues have also had a actively excludes the use of FOBT around screening procedure at this age. I would New Zealand as a referral reason for colo- suggest that as a profession, it is not appro- noscopy?4 Is it ethical that they knowingly priate for us to screen ourselves for this allow the premature death of hundreds of disease, and not try and help establish a New Zealanders each year to bowel cancer? national screening programme similar to What would happen if every family who our own personal practice. With sadness, I lost a loved one to bowel cancer challenged accept that we live in a society where “The the Ministry of Health for not acting to rich stay healthy. The sick stay poor”.7 detect early bowel cancer? If we, as a nation, viewed over 500 Bowel cancer screening has been proven unnecessary premature annual deaths as since it was first published 22 years ago,5,6 a war, then history would demonstrate so do we really need to reinvent the wheel that New Zealand would send our most

talented youths into battle to fight this on Top sports coaches and team members our behalf. I suggest that we now view depart when their national team cannot our fight against this occult disease as a deliver winning results. Therefore, is it battle, and train appropriate warriors to time for those who partook in the 1998 fight on our nation’s behalf. Instead of Working Party on Population Screening for posing with All Blacks in their dressing Colorectal Cancer,8 and subsequent review,9 room, I suggest that our Prime Minister who still work with the Ministry of Health should try and recruit retiring professional to now hand on the baton? We can only see rugby players (with a courageous heart, further because we stand on the shoulders of proven trainability and world beating giants, but I do not accept that colonoscopy hand-eye coordination skills) and get them resources are our limiting factor—the real into the screening flexible sigmoidoscopy limiting factor is the lack of leadership to work force. If Richie McCaw can take promote and fund bowel cancer screening. one year to learn to fly a helicopter, then Sir, could you please convey to Prime how long would it take him to learn the Minister John Key that we are many years art of screening flexible sigmoidoscopy? overdue for a new innovative bowel cancer We should invite him and his sporting screening strategy. He needs to take his colleagues, both men and women, to head out of the sand and act responsibly join us at the coalface to prevent the now to prevent the continuing unnecessary unnecessary premature death of too many premature demise of over 500 New in our national clan. Zealanders each year.

Competing interests: Mr Hingston is an experienced colonoscopist who derives income from performing colonoscopy. Author information: Guy Hingston, Department of Surgery and Anaesthesia, University of Otago, Wellington Corresponding author: Guy Hingston, Department of Surgery and Anaesthesia, University of Otago, Wellington [email protected] URL: www.nzma.org.nz/journal/read-the-journal/all-issues/2010-2019/2015/vol-128-no-1426-4- december-2015/6755

REFERENCES: 1. Guy Hingston. We don’t multicentre randomised 615. doi.10.1001/ need a pilot bowel cancer controlled trial. Lancet jama.2014.8266 screening programme: 2010; 375: 1624–33. 4. https://www.health. Please just get on with Published Online April govt.nz/system/files/ it! NZMJ 25 June 2010, 28, 2010 DOI:10.1016/ documents/pages/refer- Vol 123 No 1317; ISSN S0140-6736(10)60551-X ral_criteria_for_direct_acc 1175 8716 URL: http:// 3. Øyvind Holme, Magnus ess_outpatient_colo- www.nzma.org.nz/ Løberg, Mette Kalager, noscopy.pdf journal/123-1317/4200/ Michael Bretthauer, 5. Mandel, J.S, Bond, J.H, 2. Wendy S Atkin, Rob Miguel A. Hernán, Eline Edwards, Ines Kralj-Hans, Aas, Tor J. Eide, Eva Church, T.R, Snover, D.C, Kate Wooldrage, Andrew Skovlund, Jørn Schneede, Bradley, G.M, Schuman, R Hart, John M A Northo- Kjell Magne Tveit, Geir L.M, Ederer, F. Reducing ver, D Max Parkin, Jane Hoff. Effect of Flexible mortality from colorectal Wardle, Stephen W Duffy, Sigmoidoscopy Screening cancer by screening for Jack Cuzick. Once-only on Colorectal Cancer fecal occult blood. Minne- flexible sigmoidoscopy Incidence and Mortality: A sota Colon Cancer Control screening in prevention Randomized Clinical Trial. Study. N.Engl.J.Med of colorectal cancer: a JAMA. 2014;312(6):606- 1993;328(19):1365–1371

6. Hewitson P, Glasziou PP, Issue 1.Art. No.: CD001216. govt.nz/system/files/ Irwig L, Towler B, Watson DOI:10.1002/14651858. documents/publications/ E. Screening for colorectal CD001216.pub2./14651858. colorectalcancer.pdf CD001216.pub2. cancer using the faecal 9. https://www.nsu. occult blood test. Hemoc- 7. http://www.u2.com/ govt.nz/system/files/ cult. Cochrane Database of lyrics/51 page/colorectal-can- Systematic Reviews 2007, 8. https://nhc.health. cer-screening-advice.pdf

Modelling of tobacco endgame interventions: a response Richard Edwards, Tony Blakely, Frederieke van der Deen

e are writing in response to the this approach is used, the authors could link letter published in the Journal by the letter to an online report that explains WAssociate Professor Laugesen and the methods and assumptions fully (an Professor Grace on the impact of tobacco tax, action that would still be good to do, given denicotinisation of cigarettes and e-cigarettes the importance of the tobacco endgame for on smoking prevalence in New Zealand.1 New Zealand). We have several points to make about this Thirdly, based on the information that letter and the findings it presents about the is provided about the methods, we have estimated impact of these interventions on several concerns about the approach used achieving the endgame target of a smoking and assumptions made. For example, it prevalence below 5% by 2025. appears that Laugesen and Grace assume Firstly, the letter lacks context. There that all of the change in smoking prevalence is no mention of the previous modelling from 2010 to 2014 was due to the annual work on forecasting trends and on tobacco tobacco tax increases. There are three main tax interventions in New Zealand,2-6 nor concerns about this assumption. First, other any comparisons with the findings from policies were in place or introduced over this work. Previous US work on modelling this time period (eg, point-of-sale display the potential impact of denicotinisation is ban, smokefree prisons and extension of also not referenced.7 The New Zealand tax smokefree areas in many jurisdictions). modelling work carried out by BODE3 in These policies plus further denormalisation particular includes a sophisticated simu- of smoking, resulting from the adoption of lation study to estimate future smoking the Smokefree 2025 may account for some prevalence trends before and after tax of the observed decline in prevalence from increases, and through to health gains, 2010-14. Second, the modelled effect of tax costs and inequality impacts.2,4,6 This work should be that over and above business-as- includes price elasticities at the heart usual trends in smoking prevalence. A third of the modelling (as is the international problem is overreliance on two points in norm), and allows for changing demo- time (2010 and 2014). If the 2010 estimate graphic structure, competing mortality was by chance high, and the 2014 by chance risks, and so on. low, then impact of tax will be overesti- Secondly, detail on the methods and mated. In addition, Laugesen and Grace model assumptions is limited. Providing seem to have simply extrapolated a change information on methods is a fundamental in tobacco consumption into an equivalent principle for communication in scientific reduction in smoking prevalence. However, journals, and perhaps even more crucial reductions in tobacco consumption are when (complex) simulation modelling tech- made up of a combination of falls in prev- niques are applied. We understand that the alence and reduced consumption among length of letters is inevitably constrained, continuing smokers. Hence, tobacco tax making it difficult to include all necessary consumption elasticities (how much methodological information. This means consumption reduces with price increases) that letters may not be the appropriate are substantially greater than prevalence mechanism for communicating the findings elasticities (ie, how much prevalence of complex modelling studies. However, if reduces due to a price increase).

Estimates for falls in tobacco prevalence as through gateway effects for youth by way of annual tax increases of 10% or or reduced quitting among dual-using larger in previous New Zealand modeling smokers. The latter could occur if the work were lower than those reported by ability to use e-cigarettes where smoking Laugesen and Grace, presumably reflecting is not allowed, or the perception that more realistic assumptions about impact. cutting down smoking with the help For example, Laugesen and Grace estimate of e-cigarettes is sufficient, resulted in an annual 15% increase in tax will result reduced motivation to quit. These effects in a prevalence of 5.1% by 2025, whereas are plausible. For example, in the UK, Cobiac et al estimate 8.2% prevalence for whilst 41% of dual-users report using the same scenario.2 Further work by the e-cigarettes to quit, another 43% report

BODE3 group is underway to examine how using e-cigarettes to cut down, but not price elasticities may change with very stop completely, and 25% report using high tobacco prices, and the estimates may them “because I want to continue smoking subsequently be amended. tobacco, and need something to deal with However, at this point in time, the situations where I cannot smoke (eg, published projections are the best (we workplaces, bars or restaurants).”9 believe) that can be done for New Zealand. The uncertainty about the net population Similarly, the estimates of the impact of e-cigarettes on smoking preva- population-level impact of e-cigarettes may lence is illustrated by the finding in some, be optimistic for several reasons. Firstly, but not all, longitudinal population-based their assumed annual quit rate through studies that e-cigarette users do not have e-cigarettes in New Zealand (if they were higher quit rates than non-users,10 and the made more widely available) seems to lack of evidence of a substantial increase in assume that all US and UK ex-smokers who the rate of decline in smoking prevalence are currently regular users of e-cigarettes or change in quit rates among smokers in gave up during the last year, gave up using jurisdictions where e-cigarette use is very e-cigarettes and, importantly, would not common like the UK.11 have given up otherwise. This assumes, There are also uncertainties about the unrealistically, that an equal number population-level impact of denicotini- of new ex-smokers as the total current sation—such as how interventions studied number of ex-smoker, e-cigarette users in experimental settings (eg, controlled will quit using e-cigarettes every year trials using denicotinised cigarettes) may subsequently. Furthermore, it assumes impact on smoking prevalence in real life. none of this group of ex-smokers only Given this degree of uncertainty, the lack started using e-cigarettes subsequent to of any sensitivity analyses, discussion of quitting (and hence did not quit through alternative scenarios or at least some infor- e-cigarette use). Finally, it assumes that mation on the degree of uncertainty in the all of these ex-smokers who quit using predicted prevalences is concerning, and e-cigarettes only quit because of their use contrasts with previous modelling work.2,4 of e-cigarettes—ie, none would have quit anyway using other means (NRT, Quitline Finally, we note that the Smokefree 2025 support etc) if e-cigarettes had not been goal was derived from a recommendation available. An example of the importance of the Māori Affairs Select Committee.12 of taking the latter into account is a recent Given the much higher smoking preva- estimate of the impact of e-cigarettes on lence among Māori in New Zealand, there prevalence (authored by a supporter of are concerns that the Smokefree 2025 goal e-cigarettes in tobacco control) which may be achieved for the overall population, used the marginal effect of quitting using whilst Māori smoking prevalence remains e-cigarettes over and above quit rates from well above 5%. We suggest, therefore, that unassisted quit attempts.8 all modelling work should present estimates The estimates of the impact of by ethnicity, as is the case in other recent e-cigarettes also ignore potential (but New Zealand modelling studies. unproven) negative impacts of e-cigarettes We believe that work like Laugesen and on reducing smoking prevalence, such Grace’s modelling the potential impact

of endgame interventions is important However, we are concerned that the and can help inform the development of estimates presented provide unjustifiably evidence-based strategies for achieving optimistic and misleadingly definitive Smokefree 2025. We are sure that the estimates of intervention effects, and hence authors are strongly committed to skew debates about how best to achieve the enhancing constructive discussion around Smokefree 2025 goal. achieving that goal. Competing interests: Nil Author information: Richard Edwards, Public Health, University of Otago, Wellington; Tony Blakely, Public Health, University of Otago, Wellington; Frederieke van der Deen, Public Health, University of Otago, Wellington. Corresponding author: Prof Richard Edwards, Public Health, University of Otago Wellington [email protected] URL: www.nzma.org.nz/journal/read-the-journal/all-issues/2010-2019/2015/vol-128-no-1426-4- december-2015/6576

REFERENCES: 1. Laugesen M, Grace ment in smoking cessation Britain. London: Action R. Excise, electronic services in New Zealand. on Smoking and Health cigarettes and nicotine Am J Public Health. (England), 2015. reduction to reduce smok- 2010; 100:1274-81. 10. Hitchman SC, Brose ing prevalence in New 6. Ikeda T, Cobiac L, Wilson LS, Brown J, Robson D, Zealand by 2025. N Z N, Carter K, Blakely T. McNeill A. Associations Med J. 2015; 128:1420. What will it take to get Between E-Cigarette 2. Cobiac LJ, Ikeda T, Nghiem to under 5% smoking Type, Frequency of Use, N, Blakely T, Wilson N. prevalence by 2025? and Quitting Smoking: Modelling the implications Modelling in a country Findings From a Longi- of regular increases in with a smokefree goal. Tob tudinal Online Panel tobacco taxation in the Control. 2013; 24:139-145 Survey in Great Britain. Nicotine Tob Res (Advance tobacco endgame. Tob 7. Tengs TO, Ahmad S, Access published April 20, Control. 2015; 24:139-45. Savage JM, Moore R, Gage 2015). Available at: http:// 3. van der Deen FS, Ikeda E. The AMA proposal ntr.oxfordjournals.org/ T, Cobiac L, Wilson N, to mandate nicotine content/early/2015/04/20/ Blakely T. Projecting reduction in cigarettes: a ntr.ntv078.short future smoking preva- simulation of the popula- 11. West J, Brown J. Latest lence to 2025 and beyond tion health impacts. Prev trends on smoking in in New Zealand using Med. 2005; 40:170-80. England from the Smoking smoking prevalence data 8. West R. Estimating the Toolkit Study Edition. from the 2013 Census. N population impact of London: University of Z Med J. 2013; 127:71-9. e-cigarettes on smoking College, London, accessed 4. Blakely T, Cobiac L, cessation and smoking September 17 2015. Cleghorn CL, et al. Health, prevalence in England. Available from: http:// Health Inequality, and London, 2015. London: www.smokinginengland. Cost Impacts of Annual University of College, info/ latest-statistics/ Increases in Tobacco London. Accessed 12. New Zealand Parliament. Tax: Multistate Life September 17 2015. Inquiry into the tobacco Table Modeling in New Available from: http:// industry in Aotearoa Zealand. PLoS Med. www.smokinginengland. and the consequences of 2015; 12:e1001856. info/sts-documents/ tobacco use for Māori. 5. Tobias MI, Cavana RY, 9. ASK UK. ASH Factsheet Report of the Māori Bloomfield A. Application on the use of electronic Affairs Select Committee. of a system dynamics cigarettes (vapourisers) Wellington: New Zealand model to inform invest- among adults in Great Parliament, 2010.

Safety and efficacy of digoxin Recent observational studies have suggested increased mortality associated with digoxin in those with heart failure and atrial fibrillation. These observational findings are in contrast with the results of the digoxin in heart failure (DIG) trial. This meta-analysis reviews data from recent observational and controlled trials. All studies published from 1960 onwards that examined comparative outcomes with digoxin and control (placebo or no treatment) were reviewed. The primary outcome was all cause mortality. Fifty-two studies contributed to the systematic review, including 621,845 patients who received treatment with digoxin or control. The researchers noted that prescription bias may have contributed to the observational trial findings, ie clinicians prescribing digoxin to patients at the highest risk. The conclusion of this meta-analysis is that digoxin should continue to be considered as a treatment option to achieve heart rate control in those with atrial fibrillation and also to avoid hospital admission in patients with heart failure. An editorial review commends the findings and observes that trials are best, ignore the rest. BMJ 2105;351:h4451 and BMJ 2015;351:h4662

Perioperative bridging anticoagulation in patients with atrial fibrillation It is uncertain whether bridging anticoagulation is necessary for patients with atrial fibrillation who need an interruption in warfarin treatment for an elective operation or other elective invasive procedure. This report concerns a randomised, double-blind, placebo-controlled trial in which, after perioperative interruption of warfarin therapy, patients were randomly assigned to receive bridging anticoagulation therapy with low-molecular-weight heparin or matched placebo administered subcutaneously twice daily for 3 days before and then 5 to 10 days after the procedure. 1,884 patients were randomised. The incidence of arterial thromboembolism was 0.4% in the no-bridging group and 0.3% in the bridging group and the incidence of major bleeding was 1.3% in the no-bridging group and 3.2% in the bridging group. The researchers conclude that forgoing bridging anticoagulation was non-inferior for the prevention of arterial thromboembolism and decreased the risk of major bleeding. N Engl J Med 2015;373:823-33

Compliance of males with stage 1 testicular germ cell tumours on an active surveillance protocol Testicular germ cell tumours (TGCT) are the commonest solid tumour in young men. Stage 1 TGCT (ie, confined to the testis) requires orchidectomy, followed by management options that include chemotherapy, radiation treatment, retroperitoneal lymph node dissection or active surveillance. Surveillance avoids the morbidity of the other treatments and all treatments have an excellent outcome. Successful surveillance requires patient compliance with rigorous follow-up. This study examines the rate of compliance in follow-up of 57 men in the surveillance programme. At median follow-up of 24 months, 81% had adequate compliance with the follow-up regimen, 12% were lost to follow up, and 16% relapsed; none between protocol visits. Methods to increase compliance are needed. The authors suggest that nurse-led or internet schemes might be helpful. Internal Medicine Journal; 2015,45:1081-84

URL: www.nzma.org.nz/journal/read-the-journal/all-issues/2010-2019/2015/vol-128-no-1426-4- december-2015/6757

Post mortem examinations a public service

Hamlet’s Soliloquy. A Fancy Morgue Sketch. The Coroner: To be or not to be?— that is the question. (Observer, 24 August 1912). Alexander Turnbull Library, Wellington, New Zealand. http://natlib.govt.nz/ records/27572970

his subject has been well treated by The opinion of a pathologist should at Dr Wilson, of the Mayo Clinic. Prob- once be available on the nature of any Tably in no department is there more growth removed at the time of operation, scope for general education than in this and in the case of death, both relatives important one. In New Zealand today, if any and practitioners should welcome a post of our valuable stock die, or an epidemic mortem examination, for the invaluable effects our sheep or horses, post mortem information that both may obtain. examinations are made and the result The fact that the public do not recognise communicated to the owners. Surely our the value of post mortem examinations, own children are as valuable an asset as and consequently ref use permission when our cattle? one is requested, is largely the fault of the In this, as in other serious diseases, the profession. A concise and straightforward medical profession will make no headway report by a skilled pathologist should be until it convinces the public of the value handed in writing to the next-of-kin after of expert bacteriological or pathological any such examination. It is not right that findings. In cancer, up to the present, we the physician alone should gain knowledge have unfortunately no bacteriology, but by such examination. It is of great impor- through the pathologist alone can reliable tance that a man should know what really statistics, reliable diagnosis in early cases, caused the death of his child. Moreover, and the true story in neglected cases be no greater stimulus is possible to a sound accurately unfolded. clinical work than the verification by post

mortem findings. The medical profession the profession and public interested in the is, or ought to be, the guardian of the public subject. Valuable information, both to the health. Anything that tends towards the profession and to the public, should thus be greater efficiency of the profession should obtained, and uniform methods of dealing be welcomed by the public if it is mindful with the disease established throughout of its dearest asset—good health. The the Dominion. Similar conferences have pathologist is, or ought to be, the auditor of frequently been held on the subject of medical accuracy. tuberculosis with excellent results, and I In conclusion, I make the following believe that the cancer problem should not suggestion: That the Council of the NZ present the same difficulties today that the Branch of the British Medical Association consumption problem did ten years ago, make arrangements for convening a cancer provided that the full forces of education conference, to which be invited the heads and scientific investigation are brought to of the Health Department, Veterinary bear on the subject in an organised manner Department, chairman and members of throughout the Dominion. Charitable Aid Boards, and all members of NZMJ December 1915

URL: www.nzma.org.nz/journal/read-the-journal/all-issues/2010-2019/2015/vol-128-no-1426-4- december-2015/6758

Selected proceedings of the APAC Forum 2015

Hidden Hospital Results: From 2009 to 2014 areas. In response to the New there was a 21% increase Zealand government’s target Tim Denison in adult acute elective and of 98% participation in ECE Programme Director, Auckland arranged presentations from by 2016, Ko Awatea and the District Health Board’s Performance 69,225 to 83,279. During this Early Learning Taskforce led Improvement Team, Auckland period the average wait time the Early Learning project to Aims: In 2009, only 63% of to be admitted from ED to an increase ECE enrolment and patients in the adult emergency inpatient ward bed reduced attendance rates in such areas department (ED) at Auckland from 8 hours (2008 & 2009) to of South Auckland. City Hospital were admitted, under 1½ hours (March 2011 Method: The project was discharged or transferred from through June 2015). Direct care structured as a collaborative ED within 6 hours of arrival. A time of nurses with patients with seven ECE centres partici- major contributor to patients increased from 34% to over pating, using the Breakthrough exceeding 6 hours in ED was the 60% on 10 wards. This equates Series Collaborative Model for inability for patients to access to 2 more hours per nurse per 8 Achieving Improvement (BTS). hospital beds (access block). The hour shift. The BTS included three learning aims of this work were to: Conclusion: Access block at sessions interspersed with • Reduce waiting time for Auckland City Hospital has been action periods. Participants were admittance from ED to an reduced through improvement trained in improvement method- inpatient ward bed initiatives facilitated by ology at the learning sessions and experienced improvement • Reduce access block by used plan, do, study, act cycles practitioners working alongside reducing waiting time for to develop and test changes healthcare workers to apply patients throughout their according to local context during Lean Six Sigma and lead hospital stay the action periods. change. Participants collected Methods: Improvement 1. Toomath R, Szecket N, Nahill weekly data on the number specialists from manufacturing, A, Denison T, Spriggs D, Lay C, of licenced places occupied banking and management et al. Medical service redesign with an enrolled child, booked consulting with experience shares the load saving 6000 bed days and improving and attended hours, and the applying Lean Six Sigma and morale. Internal Medicine percentage of available capacity leading change formed a Journal. 2014; 44(8): 785-790. used. Data was entered into an Performance Improvement Excel spreadsheet and collated Team (PIT) at Auckland DHB. Improving early weekly by Ko Awatea. The PIT’s key initiatives childhood outcomes Results: The aggregated included: training 23 healthcare (Auckland, New Zealand) staff in Lean Six Sigma Green median enrolment rate rose Belt Training to lead their own Jilly Tyler1, Monique Davies2 from 76% in January 2014 to projects; applying Lean Six 1Early Learning Taskforce Project 89% in June 2014. Average Sigma principles to medical Lead, Ministry of Education; 2Collab- attendance rose from 12.5 rosters to evenly distribute orative Project Manager, Ko Awatea hours each week per child in patients per doctor and align Background and aims: January 2014 to 18.6 hours each shift times with patient presen- Participation in high quality week per child in June 2014. tation patterns1; and delivering early childhood education (ECE) Conclusion: Using BTS a communications programme is associated with improved and Model for Improvement to engage clinicians in identi- cognitive outcomes at school methodology enabled partici- fying and eliminating causes and the development of charac- pating ECE centres to increase of waiting. Finally, the PIT teristics that support learning.1 enrolment and attendance rates facilitated over 300 ward staff Children who do not participate within existing resources. Staff to improve efficiency of tasks in ECE, or who attend regu- in ECE centres developed the away from the bedside so that larly for less than one year, capability to develop and test nurses could spend more time are disproportionately from changes for improvement and to with patients. socio-economically vulnerable understand their effectiveness.

1. Mitchell L, Wylie C, Carr M. • Care coordination implemented more broadly Outcomes of early childhood working with primary and and successfully through education: literature review. [Wellington, New Zealand]: secondary teams to facil- considering: the strategic Ministry of Education; 2008. itate patient transitions context; key success factors and to a lower level of care, if barriers to implementation; Care integration and appropriate. existing systems which it can be coordination: Learning • Cross-organisation built on. from a Qatari pilot pathways to support effi- Methods: Phase 1: Eighteen cient and seamless care semi-structured interviews Ihab Seoudi1*, Nicola Dymond1, integration for patients with internationally recognised Eden Magana1, Enaam Abu Donia1 transitioning between experts including physicians, ,Vicky Scruby1, Mahmoud Al Raisi1 providers. nurses, researchers, religious 3 1 , Saleh Al Marri , Marcos Alman , Results: Surveys revealed leaders and lawyers. Cathy Waters2, Flora Asuncion2, that 85% of patients did not Phase 2: Four qualitative Steve Phoenix1, Alison Robertson1, have a formal care coordinator case studies of US health Nabeel Shaath1, Ayman Tardi1, and expressed the need for systems identified as leaders Samya Al Abdulla2, Bilal Khan2, Kerri better care coordination and in end-of-life care and advance Mackenzie1, Robert Moorhead3 and integration of services across care planning. Sixty semi- Steven Archer3 providers. structured interviews were 1 Hamad Medical Corporation, conducted to support the case 2 Inpatient assessments Qatar; Primary Health Care studies, mainly with physicians, Corporation, Qatar; 3Supreme revealed that 46 patients (21%) nurses, senior leaders and Council of Health, Qatar had clinical needs that could have been more appropriately change managers. Background and aims: delivered at lower levels of There is increasing pressure Notes and recordings from care. The top 3 reasons for this on the health system in Qatar, interviews were transcribed were: which is creating a strong and coded thematically using driver for integrated and coor- • lack of a defined care Atlas.ti software. Participants dinated care. As part of Qatar’s continuum confirmed themes identified. National Continuing Care • limited capacity or absence Results: The research Design Strategy, a pilot was of an alternate care setting identified seven foundational elements for the launched to assess care coordi- • limited home and implementation of advance nation and integration across community services the healthcare system. care planning: Workshops revealed that 1. Strategic fit: a strategic Methods: Lean Six-Sigma the implementation of cross- commitment to patient- was used as the project organisational pathways and family-centred care management and improvement was hindered by absence of which supports and methodology in a 6-month pilot system-wide care coordination as it offers a structured rigorous and system navigation aligns with advance care data driven approach. The pilot functions and a technology planning. was based on the following key infrastructure to support 2. Cultural change: this components: patient referral and transition should first be promoted • Patient and healthcare between providers. in the community, so there provider engagement Conclusion: In order for care is an increased level of through focus groups, coordination and integration comfort with the concepts workshops and surveys. initiatives to be successful in of end-of-life and death. • Healthcare provider Qatar, a strategic framework Second, promotion should partnership between and mandate for system-wide occur within health primary and secondary implementation are needed. systems so that death is care providers in Al Wakra Approached in this manner, not seen as a failure and municipality. there is a potential to rebalance the system recognises the existence of a range of • Concurrent patient demand across the healthcare available end-of-life care assessment of 220 inpa- system and provide patient- options. tients to determine if they centric care. were placed at the most 3. Senior leadership: senior appropriate level of care for Values-based health: leaders providing active, their clinical condition and Improving care towards engaged and visible if not, identify the reasons the end of life leadership. for that and recommend 4. Dedicated resources: first, a more appropriate level Helen Mason, MBA, RN, Bay of Plenty quality improvement/ of care. Case managers District Health Board, Institute for change management used the Managing Care Healthcare Improvement resources for imple- Appropriately for Patients Introduction: This research mentation and second, (MCAP) tool to conduct the sought to identify how dedicated resources for the assessments. advance care planning can be actual delivery of advance

care planning services Methods: ‘Say Ahh’ is a determinants of health. This should be provided. school-based throat swabbing approach has resulted in this 5. Embed in workflow: and treatment programme in programme being successful. establishing advance care nine low decile schools covering planning as an organisa- a population of approximately Achieving a tional expectation, and 1,800 children. The programme 7-day service making it easy to do. is provided by health workers (school nurses or kaiawhina). Jane Evans 6. Pre-existing infrastructure: An opt-in consent process Eastern Health an electronic medical was followed for all children record, preferably Background and aims: attending those schools aged available across the Eastern Health has three acute 5-14 years. Health workers continuum of care; existing hospital sites and provides visited each school 2-3 times physician education on inpatient services to over 11,000 per week to assess children communication skills; General Medicine patients with sore throats. A throat and established standard per annum. The aim of this swab was taken and antibiotic organisational approaches improvement work was to treatment offered to those to change management. reduce the variation in LOS with a positive result. House- ranging from 3.6–11.6 days 7. Amend to local context: holds of children with positive to bring it in line with major adapting and amending sore throats were assessed metropolitan health service known approaches to to identify contacts who may peers who had a LOS of around advance care planning require throat swabbing. 4.0 days. to the health system’s Affected families were offered particular context. referral to a social worker Methods: A Rapid Improvement Event (RIE), Discussion: Advance care and a comprehensive housing attended by 30 clinical staff and planning is more likely to be assessment was undertaken, 4 patients, was held in mid-2012 implemented successfully if followed by housing insulation to collectively develop and foundational elements are in repairs as required. Assistance implement a model of care that place. Dedicated resources, was also offered for access to delivered a consistent service ongoing effort and change Work and Income, Housing NZ to patients regardless of unit or management are essential. In and other Government organi- site, 7 days per week. addition, cultural change in sations. Local Non-Government attitudes towards end-of-life Organisations and charities The RIE identified three key in communities and health supported the programme by areas for change which were systems is required. offering household items for implemented on 13th June families, such as curtains, beds 2013: Eradicating rheumatic and linen. 1. Consultant led ward based fever from Hawke’s Bay Results: The total number of multidisciplinary teams throat swabs taken in the ‘Say based on a 7 day week Dr Caroline McElnay1, Dr Janine Ahh’ programme for the period 2. Agreed standard daily Stevens2, Dr Liffey Rimmer3, Julia studied (Oct 10-Sept 14) was work for all team members Haydon-Carr4, Susan Stewart5 13,311, with a positivity rate (7 days per week) 1 Director of Population Health, between 11.4% and 17.5% and incorporating: Hawke’s Bay District Health Board, no reduction in the positivity 2 3. Electronic Patient Journey Public Health Medicine Registrar, rate seen over the period. The Mid-Central DHB, Palmerston Boards to make patient overall rate of ARF in Hawkes North, 3General Practitioner & flow visible to all staff and Bay has decreased from 7.0 Clinical Lead, Hastings, 4Population all wards Health Strategist, Population cases per 100,000 in 2010 to 1.3 Health Services HBDHB, 5Team cases in 2014. In the five years To support the new structure, Leader Public Health Nurses, before implementation, the physician hours were altered to Child health Team, HBDHB annual incidence of ARF among ensure the ward based teams were in place on the weekends. Background and aims: children aged 5-14 at ‘Say Ahh’ Allied Health staff and Phar- Acute rheumatic fever (ARF) schools was 221.1 per 100,000. macists also increased their is a serious but preventable In the four years after imple- hours of treatment provided to disease. Between July 2006 mentation, the rate was 76.2 per general medicine patients on and June 2009, there were 26 100,000. This is a statistically the weekends. admissions for ARF (5.8 per significant reduction of 66% 100,000 people) at Hawke’s Bay (p=0.02). Measures included: District Health Board. Twelve Conclusion: The programme • Average Length of Stay of these cases occurred in the delivered a high quality equi- (ALOS) suburb of Flaxmere, Hastings. table service that reached those • Activity data and in The ‘Say Ahh’ rheumatic fever children most at-risk from particular, the number of programme was implemented in ARF and continually linked patients treated 2010 to reduce the incidence of the screening and treatment ARF in children aged 5-14 years programme with actions • Bed-days attending Flaxmere schools. to improve the underlying • Bed-day savings

• mortality Background and aims: As Nationally, the proportion of • Staff survey 9 months post part of the English Department patients receiving ‘harm free’ implementation of Health’s (DH) Quality, care has increased from 91.8% Innovation, Productivity and in July 2012 to 93.9% in May Results: Compared to the Prevention (QIPP) ‘Safe Care’ 2015. In the national data, from previous year, after 12 months, programme in 2010, 161 the first to the last median, this the new model of care resulted in: organisations came together represents a decrease in harm • ALOS reduction of 0.9 days to improve four high volume from 8.2% to 6.1%, or a 25% from 6.12 to 5.24 days harms: pressure ulcers, falls, decrease in the presence of • 18% reduction in mortality urinary tract infection (UTI) harm. from 4.61 to 3.78 deaths in patients with catheters and Conclusion: PDSA method- per 100 separations venous thromboembolism ology can be used to develop • 476 more separations from (VTE), which are estimated to a measurement tool for use at 10,747 to 11,212 patients collectively affect over 200,000 point of care in a variety of care 1 per annum patients per year in the English settings. The ‘harm free’ care NHS. • 7,015 bed days saved measure engaged teams and across the three sites Methods: A point of care detected change over time at a measurement tool, the NHS local and national level. • 19 less beds across the Safety Thermometer (ST), was Eastern Health General 1. Harm Free Care. “Harm Free developed to measure these Care”. www.harmfreecare.org. Medicine Service. four harms according to seven n.d. Accessed 2 April 2013. • Reduction in variation design principles:2 that the tool 2. Power M, Stewart K, Broth- in the LOS between units would have clinically valid erton A. “What is the NHS reduced to an average Safety Thermometer?” Clinical definitions, be efficient, be used Risk. 2012; 18: 163-169. 4.0–6.5 days wherever the patient is treated, 3. The W. Edwards Deming Qualitatively, physicians provide immediate data over Institute. “The PDSA Cycle.” felt that the new model time, measure all harm experi- The W. Edwards Deming improved the quality of care enced by the patient regardless Institute. 2015. Accessed October 20, 2015. provided to patients, improved of avoidability, measure harm communication between the at the patient level enabling a Post orthopaedic allied multidisciplinary team, and composite measure of ‘harm sped up the senior decision free’ care (the absence of the health assistant (AHA): making process. four harms) and be easy to Hip and happening aggregate. Weekend discharge perfor- Annette Davis, Stuart Cavill, Gary mance improved considerably Various iterations were Hannah especially on Sundays with total produced using Plan, Do, Study, Allied Health Workforce weekend discharges increasing 3 Act (PDSA) cycles working Innovation Strategy Education from 710 to 1,098 (an increase with frontline teams and & Research (WISER) Unit, Phys- of 54.6%) in the 6 months prior content experts. Data were iotherapy Department Monash to the new model (Jan-Jun 2013) collected at the point of care Health, Clayton, Australia to the second 6 month of the through conversation with the Aim: Health care globally new model (Jan-Jun 2014). In patient, patient assessment and is experiencing funding the same period there were an reviewing patient notes. and resourcing issues with extra 222 Sunday discharges, an In 2012, all NHS organisations increasing numbers and increase of 72.8%. were incentivised through a complexity of patients to Conclusion: This work has Commissioning for Quality and manage.1 It is necessary to resulted in increased standard- Innovation goal to use the ST on review service and disci- isation on work flow, reduced 100% of patients on one day per pline structures to provide LOS and the perception on month. Data were analysed and cost effective patient care increased decision making. publicly available using run without compromising quality All changes to clinical staffing charts to describe special cause and safety. One strategy is to with the new model were cost variation, and Pareto charts and optimise and upskill the allied neutral with no compromise to funnel plots to understand vari- health assistant (AHA) work- teaching and training. ation. Data were used locally force to practice at the higher to set improvement goals and end of their scope. This project Taking a temperature measure improvement. aimed to assess the cost effec- check on safety learning Results: The number of organi- tiveness and efficacy of the from measuring sations submitting data increased Monash Health Acute Ortho- improvement at from 428 (July 2012) to 822 (May paedic AHA role. scale with the NHS 2015) with 7,861,432 patients Methods: This project safety thermometer surveyed, an average of 201,575 required the reallocation of an per month. Settings where data AHA to fill a physiotherapist Abigail Harrison1, Dave Shackley2 are collected include acute and (PT) vacancy due to financial 1Haelo, United Kingdom, 2Salford Royal community hospitals, care homes and resource challenges. A NHS Foundation Trust, United Kingdom and patients’ own homes. range of productivity and

quality data was collected results fail to spread or sustain. Discussion: Teams 12 months prior to AHA This study aims to examine embarking on improvement recruitment and 12 months the factors that need to be in projects should be aware from after recruitment, including the place to achieve spread and the outset of the factors that number of patients attending sustainability. support spread and sustain- therapy, number of therapy Methods: The study began ability and incorporate these sessions, and length of stay, with a literature review via into the project. incidents and discharge Access Portal, Google Scholar, 1. Gustafson D, Maher L, Evans A. destination. British National Health Service Medline and Ovid. Sustainability Model, 2003. Results: The staff cost saving In October 2014, nine of a Grade 3 AHA as compared health-related organisations in Transformed access to a Grade 2 PT approximately England and Scotland which $27,340/annum. Comparing pre- to appointments: had demonstrated success A study with eight and post- role implementation in spreading and sustaining data, the number of separa- improvement initiatives healthcare institutions tions increased from 131 to 181 were visited under a Hospital within Singapore (38%); the number of occasions Alliance for Research Collab- Health Services of service (OOS) increased from oration (HARC) scholarship. 542 to 724 (34%); total time for 1 2 Interviews were conducted Tan A , Lee CE (Main Author), Wai all OOS increase from 19,035 to 4 3 1 2 1 with 40 individuals and C , Goh A , Toi J , Tan B , Koh S , Goh 30095 minutes (58%) with the L6, Tse D6, Lui NL1, Tan J5, Voon CL6, meetings held with a total average OOS time increasing Choo CK2 of 190 clinicians and quality from 35.12 to 41.57 minutes 1 improvement specialists at all Singapore General Hospital (SGH); (18%). Total length of stay 2 3 levels of these organisations to Singapore Health Services; KK reduced from 126 to 91 days Women’s and Children’s Hospital learn what factors they iden- (38%); average length of stay for (KKH); 4Singapore National tified as essential for spread 5 non-complex hips reduced from Eye Centre (SNEC); National and sustainability. 6.2 to 5.3 days and non-complex Heart Centre Singapore (NHCS); 6 knees reduced from 5.4 to Results: A thematic analysis SingHealth Polyclinics (SHP) 4.8 days. The cost saving, per identified factors that need to Introduction: Singapore patient per day on this ward is be in place to achieve spread Health Services (SingHealth) approximately $615. Discharge and sustainability. accounts for almost 50% of all to home increased from 83 to Enablers to spread: outpatient services delivered in Singapore public healthcare. 113 (40%). There was no change • Strong leadership at exec- Before Sept 2013, many Sing- in adverse clinical incidents utive and project team reported or complaints or Health institutions experienced levels. compliments during the time high call volumes and aban- period investigated. • Culture, people feeling able doned calls in our call centres. and supported to work in Patients from our primary Conclusion: Reallocation of new ways working healthcare network experi- lower acuity ambulation and enced long waiting times (~25 therapy tasks from a PT to an • Significant levels of mins) for referrals to specialist AHA, utilising the ‘right person ownership, time, will, outpatient clinics. In addition, right job’ philosophy, has courage and effort. confusion arising from incon- resulted in increased therapy • Strong belief in self and the sistent rules for booking and at less cost without compro- team. coordinating appointments mising quality of care. This role • Effective communication across SingHealth institutions has been demonstrated to be a to all stakeholders about resulted in high internal call patient centred, cost effective the problem being solved. volumes (~67k calls per year or and quality service provision 8.5% of annual call volume) and model. Enablers to sustaining: increased workload. 1. Unlocking skills in hospitals: • Staff stability. better jobs, more care. • Strong leadership at exec- Aim: To improve the acces- Stephen Duckett and Peter utive and project team level. sibility of the appointment Breadon. Grattan Institute, system for patients and staff. Australia, April 2014. • Availability of tools (such as the British NHS Methods: Gemba walks Spread and Spread and Sustainability were used to observe issues; sustainability factors Model1) and training about we developed process maps, sustainability. and identified bottlenecks and for successful waste. Detailed, large-scale healthcare • Staff understand how surveys were conducted for improvement to effectively sustain patients, caregivers and ground improvements staff (500 staff surveyed, 83% Cathy Vinters • Effective communi- response rate). Insights gained Clinical Excellence Commission (CEC) cation to all stakeholders were used to develop and test Background and aims: about the benefits of the changes using plan, do, study, Many projects with excellent improvement. act cycles.

Clinical and administrative Eastern Health tested 10.5 on Ward 2 during the pilot. staff from primary and tertiary whether improved compliance Critical success factors iden- healthcare institutions in with the falls performance tified were: SingHealth worked to improve standard would reduce the inci- • Ownership at the local appropriateness of refer- dence of falls among inpatients. level rals,-develop electronic referral Methods: The project applied • Data supporting the protocols, and redesign the action learning methodology improvement workflow. Appropriateness based on the performance was tracked electronically and excellence approach on two • Integrate the project into primary care physicians and inpatient wards. The approach standard daily and weekly specialists met regularly to includes three steps: work (e.g. weekly audits discuss inappropriate referrals. and meetings) Specialists provided Continuing 1. Agree the performance • High presence of senior Medical Education sessions standard. leadership share knowledge to make better 2. Monitor performance referral decisions. against the standard. • Staff opportunity to voice views and suggestions Staff across several Sing- 3. Implement improvement Health institutions were initiatives to address any • Partnering with cross-trained and empowered performance gaps. consumers. to book appointments across The methodology: Conclusion: The results institutions and disciplines. of the pilot project demon- • A ‘Rapid Improvement Appointment rules were strated a correlation between Event’ with staff and improved compliance with streamlined. Internet and consumers from two pilot mobile appointment platforms the performance and reducing wards identifying ‘barriers’ the rate of fall. The quali- were redesigned to be more to compliance with the user-centric. tative analysis also identified performance standards. a number of critical success Results: The percentage of • Plan, do, study, act (PDSA) factors and barriers to correctly fast-tracked referrals cycles to test and learn achieving compliance with the improved from 36% to 89% from interventions performance standard. This for Ortho (non-spine) urgent designed and implemented information has informed referrals. during the project. ongoing implementation of the Before intervention, primary • Leadership huddles were approach across all inpatient care clinic staff were not conducted throughout the areas at Eastern Health and a trained and given access to project to determine and sustained focus on reducing book into specialist care insti- address the human factors patient falls. tutions’ appointments. Now, preventing compliance. they can now book most of the Turning haggis into • Three focus groups (one referral appointments for 5 pavlova … Scotland specialist care institutions. for each pilot ward which included leadership and to New Zealand Abandoned calls have also front line staff and one safety in practice dropped from as high as 37% in for the project leadership FY2013, to 6% in FY2014 with team) were conducted Dr Vikas Seth, Dr Campbell Brebner, no extra staffing. at the conclusion of the Monique Davies, Andrew Jones, Ian Hutchby Conclusion: These interim project. results achieved through Counties Manukau Health Compliance with the perfor- inter-disciplinary and cross mance standard was monitored Background and aims: institutional workgroups and through an audit program. Falls The Scottish Patient Safety platforms are encouraging for rates were monitored through Programme in Primary Care SingHealth’s efforts in trans- the incident reporting system. was established in April 2013 forming patient service. Weekly performance reports to reduce the number of were provided to the Executive. avoidable harm events from Falls prevention is healthcare delivered in primary better than a cure Focus group information and care in Scotland. The Scottish data from the PDSA cycles was programme aimed for 95% of Gayle Smith analysed at the conclusion of primary care clinical teams Executive Director, Quality, Planning the pilot project to identify the to achieve reliability in three and Innovation, Eastern Health critical success factors. high-risk areas (Warfarin, Background and aims: Patient Results: Compliance with the prescribing and monitoring fall is one of the top four clinical falls standard increased from of Disease-modifying anti safety incidents at Eastern 74% in 2013 to 98% on Ward rheumatic drugs (DMARDS), Health. Analysis of the contrib- 1and from 78% to 99% on Ward and medicines reconciliation) utory factors identified failure to 2 during the 14-week pilot. The by 2016. Safety in Practice practice in accordance with the rate of falls per 1000 bed days customised the Scottish falls performance standard as the reduced from 12.85 in 2013 to programme for New Zealand most significant cause. 12.57 on Ward 1 and 16.9 to general practice (GP). GP

teams aimed to achieve safer Conclusion: Safety in that make up the index. These Warfarin management, medica- Practice improved compliance sub-scales act as a diagnostic tions reconciliation, and results with best practice among tool identifying areas where the handling. participating general prac- organisation might focus efforts Methods: We developed a tices in warfarin management, to enhance levels of measured collaborative of 23 general prac- medication reconciliation and medical engagement. The data tices to implement a primary laboratory test results handling. are collected from each participating organisation and then care trigger tool (structured combined to provide a cumu- case notes review), safety Engage to perform: lative normative dataset. climate surveys and a monthly The role of doctors audit of three care bundles. in high performing Results: UK findings: A large The concept of utilising a organisations number of significant correla- care bundle audit in the three tions were observed between chosen high risk areas is to Paul W. Long the medical engagement improve reliability in imple- Australian Institute of Health data and the set of perfor- menting best practice. Innovation, Spurgeon, P. C. mance measures (Care Quality Warwick Medical School Commission-2013). Organ- • Warfarin prescribing and isations with high levels of Background and aims: monitoring medical engagement performed Medical engagement is “the • Medication reconciliation well on the external indicators. active and positive contribution following discharge from Organisations with low levels of doctors, within their normal hospital were usually underperforming working roles, to maintaining in other areas. • Laboratory results and enhancing the performance handling systems of the organisation, which itself The relationships included a In Year 2, a fourth audit recognises this commitment, wide range of indicators, from bundle for safer opioid in supporting and encouraging clinical performance, financial prescribing was added. high quality care”. The aim management, safety indicators, of this study was to assess the patient experience and overall Practices kept individual data quality standards. to track their own progress, level of medical engagement as well as submitting their in health service organisations We found a strong association between these results bundle results for collation and (HSO) in the UK, Australia and and previously published analysis at a campaign level to New Zealand (ANZ); under- data (2008). This suggests that track overall progress. stand the relationship between medical engagement and organ- medical engagement has a Results: The monthly isational performance, and sustained probable causal link audit results showed an compare results between the to organisational performance. overall composite measure countries. Australasian results: We improvement in all three now have data from 12 sites patient safety bundles areas. Methods: All medical staff at 85 NHS Trusts in the UK and and over 2,100 doctors have Warfarin Management: 12 health service organisations completed the survey, thereby Improvement in compliance to in ANZ were surveyed, using establishing an ANZ norm. best practice from 10% in April a medical engagement scale The data reveal relatively 2014 to 74% in February 2015. (MES) instrument. The response lower levels of engagement Medication reconciliation rate at most sites was between expected at the 12 ANZ sites following discharge from 25% and 30%. The MES was based on the UK norms. hospital: Improvement in developed with a sample of Data collected at four New compliance to best practice over 20,000 NHS staff, good reli- South Wales HSO have also from 15% in April 2014 to 62% ability (0.7 to 0.93) established been analysed and compared in Feb 2015. for the sub-scales (Working to the ANZ norm. The profiles Practice test result handling: in an Open Culture, Having of medical engagement vary Improvement in compliance Purpose and Direction, Feeling at the sites and also across the from 60% in April 2014 to 90% Valued and Empowered) and MES scales and sub-scales. in February 2015. validity. This is likely to be replicated Practices self-completed The MES consists of 30 items when further analysis of the remaining eight sites is the safety climate survey and and is administered via a completed. the primary care trigger tool website link. It includes organ- (structured case review) and isational identifiers such as Conclusion: Further inves- composite group data was division, role and length of time tigation is required to explore not collected. Practice-based working in the organisation. how far the pattern of linkage meetings allowed reflection on The resulting analysis between MES and organisa- data collected and identification provides an overall index of tional performance established of changes to be tested within Medical Engagement levels as in UK is also true of ANZ data. the practice. well as scores on the sub-scales URL: www.nzma.org.nz/journal/read-the-journal/all-issues/2010-2019/2015/vol-128-no-1426-4- december-2015/6759 NZMJ 4 December 2015, Vol 128 No 1426 ISSN 1175-8716 © NZMA 128 www.nzma.org.nz/journal