A PILOT STUDY EXAMINING CRITERIA USED TO SELECT DRUGS FOR HOSPITAL, PROVINCIAL AND NATIONAL FORMULARIES Robertson J, Newby DA, Pillay T, Walkom EJ The University of Newcastle, Australia WHO Collaborating Centre for Training in Pharmacoeconomics and Rational Pharmacotherapy

Abstract A Pilot Study Examining Criteria Used to Select Drugs for Hospital, Provincial and National Formularies Robertson J, Newby DA, Pillay T, Walkom EJ the University of Newcastle, Australia Problem Statement: Little information is available on the selection criteria used to choose drugs for formularies at different levels of health care systems. Objectives: To compare and contrast the importance of 22 criteria in selecting drugs for inclusion in hospital, provincial and national drug lists in a number of emerging and developed countries. Design: Questionnaire survey of members of the Australian Pharmaceutical Benefits Advisory Committee (PBAC) and Economics Sub-Committee (ESC) (national level), 6 provincial drug selection Committees in South Africa and participants in an international short course on pharmacoeconomics who reported >1 year involvement in drug selection at hospital, area or national level (emerging and developed economies). Outcome Measures: Criteria were categorised as “always or almost always important”, “sometimes important” or “never or almost never important” for drug selection decisions and grouped as clinical factors, cost factors, drug characteristics and supply issues, and other factors influencing decision-making. Results: Responses were obtained from 11 PBAC/ESC members, 72 members of South African provincial drug committees and 22 short course participants. Clinical factors were always or almost always important at all levels of decision-making and dominant issues in drug selection. Quality of life considerations were less important than drug efficacy and safety. Cost factors were important at all levels, particularly the likely total cost (budgetary impact). Cost-effectiveness is mandated for Australian PBAC decision-making, and the potential for non cost-effective use was also important. Cost-effectiveness was reported as always or almost always important for around 90% of South African provincial decision makers, although the extent to which this is formally measured and considered is unclear. There were differences in the reported importance of cost-offsets for decision-making – potential savings to health care budgets through shorter hospital stays and reduced nursing time. The availability of alternative treatment options was an important criterion for decision-making; the age of the patients and whether or not the patient may have contributed to the development of the disease were less important factors in drug selection decisions. The importance of drug characteristics and supply-related issues varied with the decision-making setting. External pressure to list a drug was sometimes important in all decision-making settings. Conclusions: While a questionnaire cannot capture the complexity of many drug selection decisions, there were a number of clinical and cost criteria that seemed to be important in all settings. Identifying the common key elements may assist in developing formal processes to support objective and rational drug selection decisions. Study Funding: Management Sciences for Health

Introduction Bodies responsible for selecting drugs for formularies or medicines lists face increasingly difficult decisions about which drugs can be funded. New drugs may offer only small benefits over existing therapies, but at higher cost. Social, ethical and political pressures will determine health priorities and may influence drug selection. A variety of information about a new therapy must be evaluated in making drug selection decisions. While clinical efficacy, safety and acquisition costs are important, these are not the only factors considered by decision-making committees. Some factors may be context specific, i.e. relevant to hospital, area or national level decision-making.

Objectives to determine the importance of 22 criteria in selecting drugs for inclusion in hospital, provincial or national drug lists to compare and contrast the importance of the criteria for drug selection by different groups of decision makers

Methods A questionnaire survey was developed and administered to members of the Australian Pharmaceutical Benefits Advisory Committee (PBAC) and its Economics Sub-Committee (ESC), national level decision makers members of six provincial drug selection committees (PTCs) in South Africa, area level decision makers, participants in a WHO-sponsored short course in pharmacoeconomics who reported >1 year involvement in drug selection at hospital, area and/or national level, developed and emerging economies Likert-scale responses were used to classify decision-making criteria as “always or almost always” important, “sometimes” important, or “never or almost never” important.

Decision-making criteria assessed: clinical factors – drug efficacy and safety, severity of disease, other treatment options available, quality of life, whether for treatment of disease or symptom relief cost factors – cost to treat one patient, numbers of patients requiring treatment, budgetary impact (total cost), cost-effectiveness, potential for non cost-effective use, allows earlier hospital discharge, reduces nursing time ‘pharmacological’ factors – likely compliance with treatment, addiction/abuse potential of the drug, generic or branded product, reputation of the pharmaceutical company ‘other’ factors – age of patients, patient’s contribution to the condition being treated, local clinical experience with the drug, already listed on other formularies, external pressure to make the drug available

Results 1 Clinical factors: Efficacy and safety were the most important factors; the availability of other treatment options was a key consideration for all groups. Quality of life was less important than efficacy and safety. Whether the drug provided treatment of disease or symptom relief was a less important criterion for all groups (Figure 1).

Figure 1 Clinical factors rated as “always or almost always” important (block colour) or “sometimes” important (shaded colour)

Results 2 Cost factors: Consideration of total costs (budget impact) and cost-effectiveness are mandated for Australian decision makers; the potential for non cost-effective use was important for these respondents. The importance of cost-effectiveness for SA respondents was surprising as there are no formal methods for using this information in decision-making. Short course participation was often prompted by the limited use of cost-effectiveness information in their own country. Cost-offsets with reduced hospital length of stay and nursing time were of slightly greater importance to SA respondents (Figure 2).

Figure 2 Cost/cost-offset factors rated as “always or almost always” important (block colour) or “sometimes” important (shaded colour)

Results 3 The importance of ‘pharmacological factors’ varied by setting. The reputation of the pharmaceutical company and whether the product was branded or generic were not important criteria for Australian respondents but ‘always or almost always’ important for 24-43% of SA and short course respondents. Compliance with therapy was ‘always or almost always’ an important criterion for 76% of SA respondents but only 38% and 27% of short course and Australian respondents. Potential for addiction or abuse of the drug was also a greater concern for SA respondents.

Figure 3 Other factors rated as “always or almost always” important (block colour) or “sometimes” important (shaded colour).

Results 4 ‘Other factors’: The age of the population being treated; whether a person’s own actions contributed to the illness and whether the drug was listed on other formularies were more important criteria for SA respondents than for Australian and short course participants. External pressure to list a drug was recognized in all settings. Clinical experience with the drug was always or sometimes important in all settings (Figure 3).

Conclusions / Implications Survey methods cannot capture the complexity of decision-making and general questions may conceal the importance of some criteria in particular circumstances. However, the identification of criteria that are important in all settings may assist in identifying standard information requirements to support objective and rational drug selection decisions.