This month, Domhnall MacAuley takes you through a systematic review about the value of doing physical examinations before sport

Abstract


Objective-To examine the medical literature regarding the effectiveness of the preparticipation physical examination (PPE) in satisfying the basic requirements for medical screening.

Design-Medline database (1966 to August 2002) search using the terms "preparticipation physical examination" and "physical examination."

Main outcome measure-The United States Preventive Services Task Force (USPSTF) description of an effective screening test was used to determine whether the studies available provide evidence that the PPE satisfies the basic requirements for medical screening.

Results-176 articles were identified. The articles did not specify the presence of a randomised control group, nor did they describe subjects who were inappropriately cleared or restricted. Therefore, false positive and false negative rates were not presented. As the accuracy of the screening test was not demonstrated, the PPE failed to satisfy one of the major requirements for an effective screening test.

Conclusion-The PPE for athletes does not satisfy the basic requirements for medical screening as described by the USPSTF.

This month's paper is "Preparticipation physical examination for athletics: a systematic review of current recommendations" by Peter J Carek and Arch Mainous III (BMJ USA 2002;2:661). To read this paper, go to this article on studentbmj.com and click on the link.

Sudden death in the gym, on the football pitch, or in the pool-they always make the headlines and are especially poignant when the victim is young. Newspapers scream for something to be done, and everyone wonders if death could have been prevented. Perhaps the problem could have been picked up if only the deceased had had a medical examination. You might be tempted to say that this is obvious-surely medical examination can pick up all the problems. Well, the answer is not that simple. We spend our entire undergraduate medical career learning about medical examination, but it might not be as good as we think. Can a medical examination help identify problems before they occur?

Medical examination before participation in sport is not part of routine practice in the United Kingdom, nor in many other countries, but it is a routine part of medical care in the United States. The preparticipation examination has been adopted, refined, revised, and endorsed by many important medical bodies. It should therefore be possible to look for evidence of its effectiveness in identifying injury, illness, or conditions that could cause problems through sport. The authors reviewed the research literature for relevant studies that could determine if clinical examination is effective.

Medical information on any particular topic is extensive and of variable quality. If asked to look at the relevant research we could read a number of journal articles or pick some papers from an electronic database, but our selection of papers using this method may not be representative and may not tell the whole story. So, we really must search the research literature in a structured way, using a search strategy that ensures we cover all the relevant papers. We can also refine this search strategy to select only those papers that measure up to certain quality markers. We could restrict our search strategy, for example, to include only randomised controlled trials. In this study, the authors have restricted their search strategy to include only relevant papers that meet certain standards of quality.

Before publishing a systematic review, the BMJ, among other journals, asks authors to complete a checklist known as QUORUM,1 drawn up by a select group of editors of medical journals. It is useful because it allows editors to appraise the quality of a systematic review, and it also guides prospective authors in writing a systematic review. As part of a special study module at medical school, you may be asked to submit a literature review-the QUORUM guidelines can help and your systematic review could even be publishable. If you wish to look for the evidence of other treatments, you might like to look at the Cochrane database (www.cochrane.org), which is a collection of randomised controlled trials.

They wanted to measure the usefulness of the physical examination in identifying problems in athletes. So, they looked at all the research on preparticipation physical examination on the Medline database to decide if it met the criteria for an appropriate screening test. In their search they used the terms "preparticipation physical examination" and "physical examination." In addition, they searched the reference lists of those papers identified for additional papers that they may have missed in the initial search. They did not retrieve all the papers at first but looked at the titles, and abstracts, and read the full paper if it was relevant. They then collated the papers, but only included those papers in their study that conformed to certain criteria-the paper must have described the population of athletes screened and provided information on completing the examination, not meeting the criteria to pass the examination, and reasons for further examination. The authors also looked for details to measure the effectiveness of the examination-was there a comparison group, how were the groups randomised, and if it was possible to calculate the sensitivity and false negative rates. But, they found lots of problems with the papers and it was often not possible to do a proper assessment.

There are well recognised criteria for measuring the effectiveness of a screening test and these have been modified by the United States Preventive Services Task Force (USPST).2 Put simply, it is important that a test can identify a problem when it is present, and does not identify problems when they are not. If asked to measure the effectiveness of a blood test, for example, we would all be familiar with the terms false positive or false negative. Similarly, physical examination, which is another form of test, can have false positives and false negatives.

They used two basic requirements of a screening test, published in the USPST guidelines:

• The test must be able to detect the target condition earlier than would be possible without screening and with sufficient accuracy to avoid producing large numbers of false positive and false negative results
• Screening for treating people with early disease should improve the likelihood of favourable health outcomes compared to treating patients when they present with signs and symptoms of the disease

The authors found 176 articles, but they did not find what they were looking for. There were many studies which described the outcomes of physical examinations on a huge number of athletes, but none reported the outcomes in a control group that did not have an examination. Similarly, those studies recorded the conditions found at examination, but they did not provide information on athletes who were cleared at examination and were subsequently found to have a problem. As a result, the authors of the review could not identify false positives and false negatives and could not measure the overall usefulness of the preparticipation physical examination.

The authors describe a more in depth evaluation of 11 of the larger studies covering 27 780 medical examinations in total. In general, the quality of the data was limited and the samples were self selected-athletes could choose to be part of the study or go to their own family doctor. Only one study gave overall participation rates. The only useful information was on the proportion who were cleared at the initial examination, those followed up, and those cleared after follow up. The review was, therefore, unable to conclude from the studies identified if the assessments could identify individuals with a life threatening or disabling condition and identified only one death-from cerebral aneurysm-in the 27 780 athletes studied.

Yes, but for an unusual reason. In most medical research we see reports of important positive findings. This study is important because of the negative findings-the lack of scientific evidence to support medical screening. It is possible, of course, that medical screening is effective but this large systematic search could not identify studies where there was enough evidence. You may argue that these studies did find some abnormalities and you would be correct. But the authors believe that these studies only confirm that physicians detect arbitrarily established medical conditions in an unknown percentage of athletes.

The discussion section of this review is stimulating and challenging. The authors list in their introduction a number of bodies that have issued guidelines on the preparticipation medical. In the discussion, they look more closely at some of those guidelines and unpick some of the advice, using the findings of their study. They point out that the list of conditions meriting exclusion from sport includes a mixture of acute, recurrent, chronic, untreated, and inadequately treated injuries and medical problems, but that the list is based on expert opinion rather than evidence, is not standardised, and varies between studies. They point out that the studies only tell us that some conditions can be identified and may need further evaluation, and that the examination may not predict certain injuries. Listing and subsequently identifying a condition is only part of the problem.

The authors point out three main problems:

• Lack of consensus regarding the threshold of abnormality
• Unavailability of data indicating the predictive value of specific abnormalities for injury
• Lack of definitive proof that corrective intervention alters outcome.

The third problem is particularly relevant when physical examination identifies conditions such as inflexibility, muscular imbalance, and abnormal posture, which are thought to predispose to injury, but for which there is no evidence that injuries are prevented by intervention aimed at correction. There is another important point which the paper does not explore in depth-the implications of excluding someone from sport and how this may affect their subsequent health and wellbeing.

The method is good, the results compelling, and the conclusion challenges common practice. It is a useful example of a quality systematic review of an important topic. It is impossible to achieve perfection and you could criticise the search strategy-limiting the database search to Medline may be one possible weakness. In addition, the search terms used are very specific and may not have identified all the possible studies for inclusion and the authors recognise this. They also recognise that there are screening methods other than simply including history and physical examination.

However, there are important lessons from this study. The first is the conclusion that the preparticipation medical examination does not satisfy the basic requirements for medical screening. This does not mean that it is ineffective, only that medical screening is neither supported nor refuted by the current medical literature-because the evidence is not there. The second lesson is that current practice may not always be based on sound evidence from the research literature. The authors, by questioning what is widespread and common practice, have been able to challenge the advice from some medical organisations. The third lesson is that the authors had a great idea, undertook a simple review of the literature with minimal resources, and published a stimulating paper in a good journal.

1. Moher D, Cook DJ, Eastwood S, Olkin I, Rennie D, Stroup DF, for the QUOROM Group. Improving the quality of reports of meta‐analyses of randomised controlled trials: the QUOROM statement. Lancet 1999;354:1896‐900.
2. United States Preventive Services Task Force. Guide to clinical preventive services. 2nd edition. Washington, DC: USPSTF, 1996. www.ahcpr.gov/clinic/cpsix.htm (accessed 15 Sep 2003).

Guidelines: We intend publishing within 24 hours all responses that contribute substantially to the topic under discussion. To avoid points that have already been made please read other people's responses before posting your own. All responses will be eligible for publication in the paper studentBMJ, providing that current appointment and place of work are given. Name and email address are required to send a response and will be published with the response.

Please note: When writing a respnse to a response, please enter the original title of the article and the author of the response you are responding to.