Queueing for coronary surgery during severe supply-demand mismatch in a Canadian referral centre: a case study of implicit rationing

Queues for in-patient surgery are commonplace in universal healthcare systems. Clinicians and hospitals usually manage these waiting lists with informal criteria for determining patient priority — a form of implicit rationing. To understand the workings of implicit rationing by queue, we took advantage of a natural experiment in the Canadian province of Ontario. Unprecedentedly severe supply-demand mismatch led to long waiting lists for coronary surgery [CABS] in Ontario during 1987-1988. The crisis was resolved by increased funding and widespread adoption of a multifactorial clinical index for patient priority that was developed by an expert panel in 1989. Thus, we audited randomly chosen charts of patients who underwent coronary angiography at four Toronto hospitals during the crisis period, and calculated urgency scores for each case based on the multifactorial index. From 413 charts, 193 eligible patients were identified who proceeded to CABS. Waiting times did correlate with urgency ratings (r = 0.42, P < 0.0001). However, mean wait from catheterization to CABS varied among hospitals by as much as 8 weeks (P < 0.0001 after controlling for urgency scores). At the hospital with shortest queues, waiting times were twice as long for patients catheterized by cardiologists off-site vs those referred by on-site practitioners (P < 0.0001, after controlling for urgency scores); a similar form of bias was found at a second coronary surgery centre (P = 0.056, after controlling for urgency scores). Over half the patients waited longer than the maximum suggested for their category by the expert panel. Thus, implicit rationing by waiting list partly reflected clinical urgency — an advantage over American-style rationing by price or insurance cover. However, if these findings are generalizable, it appears that explicit queue-forming criteria, audits for institution-specific referral biases, and mechanisms to redistribute patients, are necessary to optimize queue-based allocation of scarce hospital services.