Skip navigation
Link to Back issues listing | Back Issue Listing with content Index | Subject Index

Cancer diagnostic assessment centres

Systematic review
Results
Comment

It is a curious fact that while individual interventions in medicine are rigorously assessed, others are not. That sentence could be taken to refer to the difference between drugs, say, which go through long and exhaustive efficacy and safety trials with complex and detailed regulatory hurdles, compared with many unconventional therapies that can be sold to patients with little or no testing.

But inside conventional medicine major interventions can be introduced without testing. These are usually top-down management changes, often introduced to 'improve' service delivery and performance where there is some perceived problem.

One such has been cancer diagnosis, with a move towards one-stop diagnostic assessment (and often treatment) centres, to better coordinate care by concentrating services, multidisciplinary consultative expertise, patient information resources and psychosocial support for patients at a difficult time. It would seem blindingly obvious that this is a better idea than any other ad-hoc arrangement. A systematic review [1] makes uncomfortable reading.

Systematic review

Multiple databases were searched for English-language articles published between 1985 and end-2002. Randomised trials, case-control studies, prospective or retrospective cohort studies were sought examining outcomes of one-stop diagnostic centres. Cancers involved included breast, lung, prostate, head and neck, or colorectal cancer, and the studies had to involve diagnostic assessment. Only full published studies were accepted.

Results

There were 20 studies (Table 1), 11 in breast cancer, three in colorectal cancer, and six in head and neck cancer. Information was reported on 18 of the 20 trials. Most were small, and 12 of the 18 reported information on fewer than 500 patients, many on fewer than 150.



Table 1: Major findings in studies of cancer diagnostic and assessment centres



Cancer Studies
Patients
Major findings
Breast 11 total
2 RCTs
6 prospective cohort
2 retrospective cohort
1 case-control
4614
1269
1084
1922
339
The two randomised trials showed no real difference in anxiety except in first few days.
Colorectal 2 total
1 prospective cohort
1 retrospective cohort
3316
3119
197
None
Head and neck 5 total
2 prospective cohort
3 retrospective cohort
427
134
293
None


Few studies examined quality of care, whether by reporting quality or accuracy of diagnosis, or patients diagnosed in a single visit, or reported clinical or economic outcomes. The two randomised trials in breast cancer hinted that patient anxiety might be lower in the first few days, but probably not thereafter.

Comment

Systematic reviews that tell us what we do not know can be more useful than those confirming what we do know. Though some work has been done on examining diagnostic assessment units in oncology, no single study has rigorously examined those managerial and clinical characteristics required to deliver a quality service.

Service changes possibly do not need randomised trials. But they do need to be evaluated, to have quality assessment programmes and checks, to be subject to audit, and even to periodic external review. None of this is new in health care. Every laboratory in every hospital does this every day to check that results are right, and to spot when something goes wrong. So do car factories.

Introducing major changes to effect improvement is good. But we need to get it right. It is a culture thing. We are unlikely to get it right first time; we will make mistakes. What we have to do is praise mistakes, for we will learn from them and be better for it. This paper [1] is a great start.

Reference:

  1. A Gagliardi et al. Evaluation of diagnostic assessment units in oncology: a systematic review. Journal of Clinical Oncology 2004 22: 1126-1135.

previous or next story