Lauren Cipriano, Bert Chesworth, Chris Anderson and Gregory Zaric.
Medical Care 46(11): 1177-1183, 2008.
Background: Total joint replacement (TJR) is clinically and cost-effective; but in many OECD countries, service provision is constrained. These constraints have triggered experimentation with a variety of strategies to improve management, with varying success. With the federal waiting time benchmark for TJR established at 26 weeks in December 2005, and with demand rising, solutions are needed to reduce waiting times and improve management of access.
Objective: To evaluate the effects of management strategies on waiting times for TJR.
Methods: Surgeons in Ontario were grouped into 25 service providing regions. Data were gathered from the Ontario Joint Replacement Registry (OJRR) on patients receiving TJR between 1999 and 2006, including number waiting by region, new patients/region/month, travel patterns, acuity and demographics. A discrete event simulation model was used to evaluate four strategies: demand reduction, priority-setting, waiting time guarantees, and common waiting time management. The simulation model started in March 2005 and progressed over 10 years, in monthly intervals, at which time determinations were made of new patients added, available surgical capacity, clinical severity and length of waiting. Each scenario was simulated 100 times. To enhance the model, sensitivity analysis was included, varying the length of initial waiting lists by +/- 20%; varying the annual increase in demand by 4% to 14%; and varying the proportion of high-priority patients between 10% and 22%.
Results: Demand reductions would have to be large and sustained to have any meaningful, long-term impact on waiting times. For instance, when demand was reduced by 25% and the number of surgeries each year grew by 16%, a minimum of 4.5 years was required to achieve 90% patients receiving surgery within six months. Clinically prioritizing patients, with more urgent patients receiving surgery before less urgent, had the net effect, on average, that 9.3% more patients received surgery within the maximum acceptable waiting time each year. Waiting time guarantees tended to reduce waiting times for targeted patients and increase it for the remaining non-targeted patients. Providing surgery within a guaranteed time to low-priority patients introduced delays for high-priority patients. Common waiting lists improved efficiency and equity and were shown to reduce regional variation in waiting times and, in the long-run, resulted in faster reductions in achieving maximum acceptable waiting times.
Limitations: This is a predictive model, at a regional level. As such, it necessarily simplifies a complex system. For instance, behavioral considerations of surgeons and patients are excluded, as are capacity limitations faced by each facility. Methodological attempts are made to mitigate problems with the accuracy and completeness of input data, by using sensitivity analysis
Conclusions: This simulation model was not aimed at precise future prediction but, rather, examined the differences expected to arise from four waiting time reduction strategies. On the one hand, only substantial reductions in demand were associated with noticeable reductions in waiting times. On the other hand, strict clinical prioritization best maximized the number of patients treated within recommended waiting times. Waiting time guarantees had little overall benefit on waiting times and any benefits came at the expense of the highest-priority patients, while common waiting lists reduced variability in waiting times and reduced waiting times in the long run.
Relevance: Though the political rhetoric may have temporarily halted, access to services continues to be a challenging problem in Canada’s publicly funded health care system. Although common waiting lists and clinical prioritization promise better management and reduction of waiting times, increasing the total number of surgeries is essential to achieving TJR within the federal benchmark of 26 weeks.
