Waiting times for cancer patients – a “slippery slope” in oncology

In his memoirs, the legendary head of the Department of Oncology (Radiumhemmet), Karolinska Hospital, Professor Jerzy Einhorn, described a chain of events during the late 1980s that led to discussions within the Swedish parliament about the state of Swedish cancer care [1]. In particular, the parliamentary discussions concerned the intolerable waiting times for patients scheduled to receive radiotherapy at Radiumhemmet. The whole affair started with Jerzy being alarmed by the increasing waiting times. He felt he had to take measures to remedy the situation by introducing stricter indications for radiotherapy. These measures gained wide media attention, were reviewed by various health care authorities, and, finally, as mentioned, were discussed by the Swedish parliament. One important outcome of these events was that Radiumhemmet was given the possibility to expand its radiotherapy resources. The alarmingly long waiting times that were felt to be intolerable, the very reason that set off this chain of events was … 2–3 weeks.

In this issue of Acta Oncologica the Danish head & neck cancer collaborative group (DAHANCA) reports the results of a survey of changes between 1992 and 2002 in the pretreatment delay for patients with squamous cell cancer of the head & neck seen at five oncology centers in Denmark [2]. They found that the median total time from first contact with the health care system to start of definitive treatment (typically radiotherapy) was longer in 2002 than in 1992 (70 versus 50 days, p < 0.001). The main reason was an increase in time from referral to radiotherapy to start of treatment (40 versus 21 days, p < 0.001). The study was inspired by a Swedish report from the Stockholm area, which also analysed pre-treatment delays for head & neck cancer patients [3]. The main impetus for that study was a major reorganisation and centralisation of the local ENT service in Stockholm during the 1990s. The median time to initiation of treatment was found to have increased over time. In contrast to the Danish study, the main reason in Stockholm was increased time used for initial diagnosis – caused by logistical problems at the university center to which all ENT cancer surgery had been centralised.

There is reason to believe that these reports of longer waiting times for cancer patients are not specific to head and neck cancer. Recent media attention in Sweden suggests that it is a growing problem also in many other, large cancer patient groups, such as breast, prostate and gastrointestinal cancer. A recent survey among the members of the Swedish breast cancer patients’ organisation, for instance, showed that about 40% of them had waited for 3 weeks or longer for their breast cancer surgery. In my experience, delays in excess of 2–3 months to initiate postoperative radiotherapy are today not uncommon in Sweden.

When these issues are discussed in the media, administrators and even physicians typically state that waiting is, of course, deplorable from a psychosocial point of view, but it is not an issue of patient safety. What is implied is that waiting, from a purely medical point of view, has no effect on outcome. It is enigmatic how such statements can be made by individuals who in other circumstances profess a belief in evidence-based medicine. In fact, the notion that waiting for definitive treatment has no effect on outcome in terms of recurrence or death due to the disease lacks scientific support. Admittedly, there are no large randomized trials that have compared longer versus shorter waiting times. Such trials would be impossible for ethical reasons. But there is other evidence that early treatment for cancer, in general, is beneficial.

Cancers are characterized by genetic instability. This leads to tumor progression characterized by a stepwise acquisition of a more malignant phenotype, which increases the risk for dissemination. Cancers have, in general, a propensity to grow which clinically results in tumor enlargement and stage migration. A larger tumor/higher tumor stage is almost invariably associated with a worse prognosis. Larger tumors are more likely to harbor hypoxic cells, which may help to explain the relationship between tumor size and decreased sensitivity to radiation therapy.

These mainly theoretical considerations are supported by clinical observations. Clinicians with an experience in, for instance, head & neck and gastrointestinal oncology are well aware of the potential for stage migration causing an initially operable tumor to become inoperable due to invasion of adjacent structures such as muscle, bone or large blood vessels.

Controlled trials of screening for cancer are in essence trials of early versus late definitive treatment. There is unequivocal evidence from such trials that early detection and treatment of both breast [4] and colorectal cancer [5] prevents death from the disease. There is today no data from trials of prostate cancer screening [6], but a randomized trial comparing early definitive surgery versus a “watch and wait” policy showed a decreased mortality in the group that was allocated to early treatment [7]. Similarly, randomized trials of adjuvant systemic therapy with an untreated control group are basically trials of early versus delayed treatment as, in most of those trials, many of the patients in the control groups were treated systemically once they recurred.

One problem with the available data is that it does not specifically tell us about the effects of waiting times in the order of a few weeks or months. Screening trials, which have shown significant benefits, have in general been those where the screening method was able to introduce lead time in the order of at least several months to some years. Similarly, in most trials of adjuvant systemic therapy, the treatment was given on average months to years earlier to those allocated to active treatment compared to those in the control group. It might be argued that effects of such considerable delays of treatment are not relevant to the current waiting times in oncology. On the other hand, it is not reasonable to postulate a threshold for the effect of delays. From a biologic point of view it is reasonable to assume that a detrimental effect is directly related to the waiting time, even if this amounts to “only” a few weeks or months.

Since the magnitude of the effect of screening as well as adjuvant therapy may be described as “moderate” (relative decreases in mortality by 10–40%), the effect of a 70 day waiting period before initiation of treatment as reported by Primdahl et al. [2] is not likely to be large, possibly not larger than an absolute survival difference of a few per cent. Differences of that size are difficult to reliably detect in observational, epidemiological studies because of random variation and confounding factors. Nevertheless, differences of that magnitude are by most patients and oncologists regarded as clinically significant and worthwhile. We should also acknowledge the possibility of individual variation. Clinical experience suggests that some patients have fast growing tumors where early treatment may be crucial, although there is still little evidence in the literature on how to identify such patients at primary diagnosis. All these circumstances should prompt us to minimize waiting times in all patients.

In summary, theoretical considerations related to tumor biology, clinical observations, and controlled trials all indicate that a delay in initiating therapy has, in general, a detrimental effect on outcome. Why then do Scandinavian health care politicians, hospital administrators, and even oncologists seem so complacent about increasing waiting times? Why are there so few outcries in the media? One reason might be that we have been on a “slippery slope” for years. Waiting times for various services that only 10–15 years ago would have been considered unacceptable have gradually become the norm. In the late 1980s a three week waiting time to initiate radiotherapy was felt to be intolerable as illustrated by the mentioned events at Radiumhemmet. Today, the median time from first contact with the health care system to start of definitive treatment among head and neck cancer patients is in the order of three months [3]. Another reason for the “slippery slope” may be that nobody is ultimately responsible for the entire continuum of care, from the patient′s first contact with the health care system to the initiation of definitive treatment. And nobody is held accountable.

How long should cancer patients need to wait to start definitive treatment for their disease? There is reason from both psychosocial and clinical points of view to initiate primary treatment as soon as possible, in most cases within days. Postoperative, adjuvant treatment should be initiated as soon as the patient has recovered from surgery, typically within some weeks. It is awkward to have to make these statements in an editorial in a cancer journal. One would have wished that the need for prompt treatment was considered indisputable, that it simply described the accepted norm and not a possible future goal. The Danish report adds to the impression that we have come too far on the “slippery slope” of waiting times.

There is an urgent need to put treatment delays on the agenda and prioritize decreased waiting times as an important goal. We must do away with the attitude that delays are inevitable and probably don't matter very much from a strictly medical point of view. We should keep in mind that there are no logistical reasons why a clinical work-up, for instance, should take more than a few days up to a week, even if it includes CT, ultrasound, endoscopic examinations, biopsies, and PET. It is mostly a matter of how the continuum of care is organised and how we prioritize, although in some instances lack of resources may contribute to the problem. The continuum of care is admittedly more complex in cancer than in many other diseases. Modern cancer care is almost invariably multidisciplinary and multiprofessional. This makes coordination difficult but nevertheless a key issue. The first measure should be to start monitoring waiting times and to identify ‘bottle-necks’ in the continuum of care. Unfortunately, of the many so-called quality registers for cancer care in Sweden, for instance, those supported by the National Board of Health, few if any have included waiting times as an important indicator of quality. Waiting times in oncology remains a curse for many cancer patients so concerted and prompt action is needed from clinicians, administrators, and politicians.

Finally, we should learn from our patients′ experiences. A man in his early sixties with a newly diagnosed prostate cancer recently described his views on cancer care at one of Sweden′s most prestigious cancer centers. After having waited for several months to start treatment for no other obvious reason than lack of coordination, he stated: “I have been through purgatory…”