Funding community medicines by exception: a descriptive epidemiological study from New Zealand

Many countries face considerable challenges in allocating resources to non-mainstream use of medicines and there is growing interest in the funding of medicines for non-mainstream uses through exceptional circumstances-type schemes.1-3 A literature review carried out as a preliminary investigation for this study indicated very little had been published internationally evaluating the operation of exceptions-type schemes.The New Zealand scheme at the time this research was undertaken was similar to some in the United Kingdom (UK).2 Mainstream pharmaceutical subsidies are provided for patients in the community at a national population level through the New Zealand Pharmaceutical Management Agency (PHARMAC), via a national community formulary (called the Pharmaceutical Schedule).4 There has been recent interest in the performance of the mainstream New Zealand scheme in relation to containing pharmaceutical costs.5The Community Exceptional Circumstances (CEC) scheme until early 2012 is New Zealand's community medicines named-patients exceptions funding scheme. It provides access to non-mainstream community pharmaceutical funding for individual patients, and is one of PHARMAC's tools for fulfilling a legislated requirement: "...in exceptional circumstances providing for subsidies for the supply of pharmaceuticals not on the pharmaceutical schedule" (New Zealand Public Health & Disability Act 2000). A risk-pool, currently of up to NZ$3 million is available from within the community pharmaceutical budget to cover the funding of such treatments.In June 2011 PHARMAC announced changes to the Exceptional Circumstances schemes following a two-stage consultation process that began in 2010. Under the new scheme, to be introduced early 2012, called Named Patient Pharmaceutical Assessment (NPPA), patients no longer need to have rare conditions to be considered for funding and there are a number of other changes. As a result the data presented in this study represents a useful historical analysis and stock-take of aspects of a scheme that has been in existence for approximately a decade. This study used national level data from the New Zealand CEC scheme. The study aimed to describe the extent and scope of the applications for funding to the CEC scheme; to analyse the PHARMAC database to assess the rates of approval/decline for CEC applications and to identify factors which are associated with successful applications.Methods All applications for CEC funding made from 1 October 2001 (when PHARMAC became responsible for administration of the scheme) to 30 September 2008 in which a decision about approval or decline available on the PHARMAC database were eligible for inclusion in the study. All other CEC application types, such as those awaiting further information or transferred to another scheme were excluded. Eighty-eight paper CEC applications not held on the database were added. The main outcome considered in the analysis was the initial application approval rate. The outcome of CEC renewal applications were not considered in detail because these applications had a very high approval rate. There were three main components to the analysis: A descriptive analysis of distributions of key variables, of overall approval rates and approval rates in relation to potential determinants; Calculation of unadjusted estimates of association (odds ratios) between potential determinants and initial approval; and Calculation of adjusted estimates of association (using multivariate logistic regression) between potential determinants and initial approval. The potential determinants of initial application outcome which were identified and included in the analysis were: Type of medicine applied for (medicine and therapeutic group); Clinical indication; Patient demographic factors including age, gender, ethnicity and socioeconomic status (SES); Type of applicant doctor classified by training and geographical location; and Application year. We used multivariate logistic regression analysis to calculate odds ratios adjusted for the following: patient gender, age, SES (deprivation index of census area unit where patient lived), and ethnicity of patient; geographical location (in Auckland or outside Auckland District Health Board (DHB) area) of applicant doctor; nature of applicant doctor (specialist, GP, general registrant (including specialist trainees and those not vocationally registered or training)) and application year. Due to the large number of indications and therapeutic groups, and the strong correlation between them it was not possible to enter these both into the logistic regression model as it would have created statistical instability. We therefore created a specialty group variable. This was a derived variable which was a combination of indication and therapeutic group grouped into categories according to the initial application approval rate. This variable was used to adjust for indication and therapeutic group within the multivariate model. Results Currently, to qualify for Community Exceptional Circumstances approval, one of the following criteria must be met: the condition must be rare; or the patient must have an unusual reaction to alternative funded treatments; or an unusual combination of circumstances applies. ‘Rare' conditions and ‘unusual' reactions are those which affect as a guide around 10 or fewer people nationally (New Zealand population approximately 4 million). Supplementary eligibility criteria include suitability of the pharmaceutical, clinical benefit, the cost effectiveness of the treatment, and, although not considered in practice now, the patient's ability to pay for the treatment. In practice, applications are made for a wide range of medications and conditions; no pharmaceutical application is not considered. Any medical practitioner can apply for CEC funding on behalf of their patients using a CEC application form or by writing to the CEC scheme. Applications are considered by a panel of six doctors which may: approve funding; decline funding; seek further information from the applicant before making a decision; or where the cost of treatment is more than $15,000 make a positive recommendation but refer the decision to PHARMAC . Applicants have the right of appeal following the CEC funding decision. Number of applications and approval rates by year—Over the 7 years from October 2001 to September 2008 there were 3234 CEC applications that were either approved or declined (Table 1). Most (2564, 79.3%) were initial applications. Overall the initial application approval rate was 16% and the renewal application approval rate was 88%. This suggested that once an approval had been given it was likely to continue to be given via renewals. The number of applications per year reduced by around two-thirds between 2001/2 and 2006/7, then increased slightly in 2007/8 (Table 1). The initial and renewal approval rates were lowest in 2001/2002 and then increased and fluctuated around the higher level. Initial approval rates were highest in 2007/8 at 34% and renewal approval rates approached 100% in 2006/7 and 2007/8. Table 1. Outcome of initial and renewal applications by application year Year Initial Renewal Grand Total Approved Declined Proportion Approved (95%CI) Approved Declined Proportion Approved (95%CI) 2001/2002 72 756 0.09 (0.07-0.11) 53 36 0.60 (0.49-0.70) 917 2002/2003 90 323 0.22 (0.18-0.26) 86 12 0.88 (0.80-0.94) 511 2003/2004 57 346 0.14 (0.11-0.18) 62 14 0.82 (0.71-0.90) 479 2004/2005 49 242 0.17 (0.13-0.22) 80 7 0.92 (0.84-0.97) 378 2005/2006 39 198 0.16 (0.12-0.22 78 6 0.93 (0.85-0.97) 321 2006/2007 47 148 0.24 (0.18-0.31) 88 1 0.99 (0.94-1.00) 284 2007/2008 66 131 0.34 (0.27-0.41) 145 2 0.99 (0.95-1.00) 344 Grand Total 420 2144 0.16 (0.15-0.18) 592 78 0.88 (0.86-0.91) 3234 Common indications and medicines—The top 20 indications accounted for over 30% of the applications (initial and renewal) over the 7-year period (Table 2). The most common three indications were osteoarthritis, depression then hypertension. Other than four transplant-related indications (approval rate 58-71%), schizophrenia (7.4%) and epilepsy (21.9%), the initial approval rate for all these common indications was less than 3%. There were 11 medicines with over 40 applications, which accounted for 32% of the initial & renewal applications over the 7 years. They were applied for under multiple indications. For six of these medicines—celecoxib, rofecoxib (both COX-2 inhibitors), venlafaxine, tramadol, clopidogrel and gabapentin—all initial applications were declined. Initial approval percents for the other five medicines most commonly applied for were cyclosporin (60%), sirolimus (48%), tacrolimus (84%) and mycophenolate (40%) (mostly for transplant indications) and fluoxetine (0.02%) (not including dispersible formulation). Approval rates by therapeutic group—Initial approval rates varied widely by type of medicine and therapeutic groups. The highest initial approval rates were for agents to treat infections and oncology agents and immunosuppressants, at 0.41 (Table 3). The lowest initial approval rate was for musculoskeletal applications at 0.01. The largest number of applications (around a quarter of the total) was for nervous system medicines, which had one of the lowest approval rates. Analysis of potential determinants of approval rates—Table 4 shows that there were significantly increased odds of initial approval among Asian, Pacific Island and Māori patients compared with European patients in the unadjusted analysis, but after adjustment for other potential confounders, all associations were non-significant except that unknown ethnicity patients had reduced odds of initial approval. The unadjusted odds ratio for females suggested there was a lower odds of initial approval among females but once adjusted for potential confounders there was no statistically significant association between initial approval and gender. There was no significant association between deprivation and initial approval rate in the unadjusted or adjusted analyses. Table 2. Twenty most common or applied for indications (initial & renewal applications) October 2001 to September 2008 Indication Total (initial & renewal) applications Percent of all (initial & renewal) applications Percent of each indication's initial applications approved Osteoarthritis 162 5.0 0.6 Depression 129 4.0 2.4 Hypertension 107 3.3 0.0 Pain 74 2.3 2.9 Transplantation of heart including failure/rejection 61 1.9 58.1 Transplantation of kidney including failure/rejection 58 1.8 61.3 Arthritis 57 1.8 0.0 Rheumatoid arthritis 52 1.6 0.0 Neuropathic pain 50 1.5 2.0 Back pain 49 1.5 2.0 Transplantation of lung including failure/rejection 45 1.4 71.4 Bipolar disorder 43 1.3 0.00