Waiting for cancer care: exploiting our national laboratory

In recent high-quality surveys of health issues, ranging from adult primary care¹ to hospital executive views² of quality of care, in 5 English-speaking countries, Canada was found to underperform on waiting and access issues. For insured individuals, we perform badly in areas such as access to family practitioners, waiting for laboratory results and joint replacement.

It seems propitious, therefore, that federal, provincial and territorial governments in the course of the September 2004 First Ministers' Agreement set objectives to reduce waiting time. Rarely has there been such a strong consensus on a single priority for reform. Perhaps this is not surprising given how well this problem is lodged in the minds of patients and administrators, as reflected in these international surveys. Waiting for cancer treatment has been the canary in the mine for the Canadian health care delivery system, because the planning myopia associated with electoral cycles has led to a lack of capacity in radiation therapy, resulting in scores of patients across the country being referred to US border cities for radiation.³

In the September 2004 agreement, each jurisdiction agreed to establish comparable indicators of access to health care professionals and to diagnostic and treatment procedures, with a report to their citizens by December 2005. The provinces and territories committed to reporting annually on their progress in making waiting-time improvements and committed to defining and reporting on benchmarks. There is certainly a range of normative opinions about the merits of waiting-time benchmarks and targeted reductions. There can be little doubt that for most cases of cancer a shorter wait is better, although the appropriateness of the rate for a given procedure is what creates the real measurement challenge.⁴ The good news is that we are getting on with reducing waiting times. We are doing this by driving the public reporting of performance measures on a number of waiting-time intervals, through a range of process improvement measures, and by increasing sheer volumes to clear the backlog of waiting patients. Ontario was the first cancer system in Canada to report in spring 2005 on a strategy-based map of measures of cancer system performance including waiting times.⁵

On the surgical side, both BC and Saskatchewan have established routine methods to track surgical waiting times. Ontario is introducing a prospective registration process for 5 priority procedures (cardiac, cancer, cataract, and hip and knee replacement procedures and magnetic resonance imaging/computed tomography). Elsewhere, billing and encounter data may represent reasonably justifiable ways to track waiting times retroactively,⁶ informed by the proviso that we must not make perfection be the enemy of the merely good. Using billing and Canadian Institute for Health Information data, Simunovic and colleagues⁷ have demonstrated that waiting times for most major cancer surgical events grew through the course of the nineties. In more recent work, it is clear that there have been continued modest increases in surgical waiting times in Ontario.⁸ Whereas there is distress associated with waiting, there is controversy regarding the state of published evidence linking adverse outcomes, including local disease recurrence and mortality, with excessive waits, as noted in the paper by Veronique Benk and colleagues in this issue of the Canadian Journal of Surgery (page 16).⁹

This article⁹ highlights important regional variation (up to 3-fold variation) in waiting intervals between surgical procedures and first radiotherapy for patients with 3 potentially curable conditions (cancer of the cervix, tonsil and larynx), as well as for a sample of women who had breast cancer resections, in the year beginning September 2001. While the range of variation in waiting intervals is significant, it is also apparent that overall waiting times for patients with potentially curable disease are considerably lower than in the case of patients who have radiotherapy for breast cancer following surgery without chemotherapy. Clearly, active clinical triage and priority setting are occurring for radiotherapy in the case of these potentially curable conditions. The paper illustrates that the overall and discrete component intervals for waiting times were shortest for the potentially curable cancers. Women who did not receive any chemotherapy between surgery and radiation therapy for breast cancer had the longest median wait of 12 weeks.

The range of regional variation in waiting times suggests that there are important and probably modifiable differences in the processes by which patients are accepted and assessed in the early stages of their treatment and in the availability of imaging and internal patient management protocols, centre by centre. Certainly there is merit in looking systematically at the relation between rates of radiotherapy use for these procedures based on the work by Barbera and colleagues, using both evidence-based and criterion-based methods to estimate the appropriateness of radiotherapy rates.¹⁰

How carefully should we attempt to codify disease-specific benchmarks and priority rankings for waiting times for patients with cancer given the current focus on this topic and the commitments by our provincial and federal governments? The Canadian Association of Radiation Oncologists (CARO) recommends 2 weeks from referral to consultation and 2 weeks from consultation to treatment or a total of 4 weeks from referral to treatment. The Department of Health in the United Kingdom has suggested a 1-month benchmark from GP referral to start of treatment for all cancers by 2005. Saskatchewan in its surgical registry has set a 6-point priority scoring system (www.sask surgery.ca/target_time frames.htm). All cancer surgery is given priority II ranking, which indicates that 95% of the surgery should be done within a 3-week target time. None of these targets are based on careful evidence summaries, because little evidence exists on which to build such targets.

In the final report of the Western Canada Wait List (WCWL) project (www.wcwl.ca), maximum acceptable waiting-time benchmarks for cataract removal and hip and knee replacement surgery were developed by compiling clinical, public and patient preferences. These stakeholders display stark differences in what is deemed an acceptable wait. Given the paucity of firm outcome information associated with various waiting times and with various clinical interventions for cancer, and while we wait for the accumulation of such data from prospective studies and natural experiments, perhaps this multistakeholder WCWL method provides a new way to benchmark maximum waiting times where some kind of consensus of provider, patient and public expectations of maximum waiting times might be used.

These initiatives have begun to affect surgeons directly. There are real concerns about a limited focus on waiting-time reductions as simply one indicator of quality. There is also legitimate concern about „cannibalization” of other domains of surgical activity within the hospital by the 4 priority areas for surgery (cardiac, cancer, cataract, hip and knee replacement procedures). Anecdotal reports of more frequent cases of empyema of the gallbladder appearing in the emergency department suggest that there may be merit in a natural experiment to examine systematically whether such displacement effects of waiting-time reduction efforts are occurring on a larger scale.

This is precisely what is exciting in the study by Benk and colleagues.⁹ The authors suggest that their study population represents a natural inception cohort among which recurrence of cancer within the anatomical part within 3 years of writing of their manuscript can be measured. This would do much to illuminate the measurement issues regarding the adverse effects of waiting for treatment for these specific disease conditions and may potentially also illuminate the population-wide effects of ramping up surgical oncology volumes. Benk and colleagues, building on their chart reviews and administrative and cancer registry data, propose to answer this question of local recurrence rate in the face of the significant variations in waiting times for treatment of potentially curable cancers. In this regard, they are surely lighting the way for investigators across Canada to create prospective cohorts with natural variation in waiting times with the aim of considering the consequences for local disease recurrence and mortality as they define performance and push for lower waiting times. While there is much national dialogue about the challenge of the recent Supreme Court judgment,¹¹ a case which turned on excessive waiting for surgery, the current national focus on waiting-time reduction and variation represents a natural laboratory in which to examine the questions of recurrence and the risk of waiting.

Competing interests: None declared.

Correspondence to: Dr. Terrence Sullivan, President and CEO, Cancer Care Ontario, 620 University Ave., 16th Fl., Toronto ON M5G 2L7; fax 416 971-5940; terry.sullivan@cancercare.on.ca