Considering the results of our limited literature searches and survey, along with our assessment of the relative importance each guideline statement contributes in producing good outcomes, we classified the AHRQ guidelines into 3 categories - withdraw; retain, append new evidence; and retain.

Table 11 lists our recommendation for each guideline, along with a succinct description of the key evidence upon which we based our decision. The new evidence we provide is not intended to be a complete recitation of all that we identified, but rather it is the evidence that most influenced our decision-making. In all, we recommended that (1) seven guidelines be completely withdrawn, because diagnostic or therapeutic recommendations that we judged contributed greatly to improving patient outcomes were obsolete or missing; (2) three guidelines be retained in circulation, as they were judged to still represent good clinical care; and (3) six guidelines be retained with new evidence appended, as the new evidence was judged to be insufficient to warrant total withdrawal of the guideline but of sufficient importance that clinicians interested in the guideline should be made aware of its existence. This new evidence could be appended to the guideline either ! electronically or via a paper insert.

Table

Table 11. Evaluation of AHRQ Clinical Practice Guidelines.

For the Determinants of Quality Mammography guideline, we had a very low response rate to our survey, and for most of the guideline statements had only a single respondent's opinion about current validity. Furthermore, we did not have within the project team the specialized knowledge of radiology needed to interpret the significance of some of the new evidence we identified. Therefore, we did not consider either our survey of expert opinion or our limited literature search to be sufficient to reach a conclusion regarding the current validity of this guideline, and have classified it as "uncertain."

For seven of the sixteen guidelines for which we made a recommendation, the decision was clear: new evidence conclusively demonstrated key diagnostic or therapeutic recommendations to be no longer valid or missing from the guideline and therefore the guideline should be withdrawn; or that there was no new important evidence and the guideline was still valid. For the other nine guidelines, more nuanced judgements were required, which we will now explain in more detail.

For both of the pain guidelines (Acute Pain Management and Management of Cancer Pain), the principal new evidence identified was the availability of newer analgesics and new routes of administration. We also found a new Joint Commission on the Accreditation of Healthcare Organizations Standard relevant to pain management that could be incorporated into the guideline. However, we judged the rest of the content of the two guidelines to be essentially valid. Rather than withdraw the entire guideline, we have judged that the information about the availability of newer analgesics and routes of administration could be appended to the existing guideline in a manner akin to a formulary update.

The principal new evidence we found the Urinary Incontinence guideline was the availability of a new antimuscarinic medication, tolteridine, to treat urge incontinence with detrusor instability. There is evidence that tolteridine is as effective as but has less cholinergic side effects than oxybutynin, the therapy recommended in the guideline. Since oxybutynin may still be a valid recommendation for some patients, and as the benefit for tolteridine is not in terms of efficacy but rather tolerability, we have judged this information to be insufficient to necessitate withdrawal of the entire urinary incontinence guideline. Rather, akin to a formulary update, the new evidence concerning tolteridine could be appended to the existing guideline.

The Sickle Cell Disease guideline was limited to screening, diagnosis, management and counseling in newborns and infants. It was released in 1993 prior to the availability of any therapies specific for sickle cell disease. It includes a guideline recommendation regarding the diagnosis and management of complications. In 1995 and 1996, data were reported demonstrating a benefit of two specific therapies: hydroxyurea and bone marrow transplantation. Although the information on screening, diagnosis, and counseling remains valid, we judged that clinicians, patients and their families would want to know about specific, potentially life-saving therapy for sickle cell disease, and therefore recommend that this guideline be withdrawn.

The existing statements in the Unstable Angina guideline were judged to be almost entirely still valid. However, a completely new class of therapy, glycoprotein IIA/ IIIB inhibitors, has emerged as an important life-saving therapeutic advance for select patients. In addition, there is evidence that low molecular weight heparin may be superior to unfractionated heparin in producing better patient outcomes, and that clopidogrel also may help prevent acute ischemic events in high risk patients. As these new treatments are all potentially or proven to be life-saving, we recommend that the guideline be withdrawn.

For the Otitis Media with Effusion guideline, the principal new evidence we identified concerns better understanding of the risks of otitis media with effusion on receptive and expressive language development. This, in turn, effects the decision about when a principal therapy, myringostomy tube placement, is indicated. Additional new evidence that we judged of lesser importance includes data concerning the efficacy of the pneumococcal vaccine at reducing episodes of otitis media with effusion, and a re-examination of the possible beneficial effect of steroid therapy in treating otitis media with effusion. As the timing of the use of a principal therapy is called into question by new evidence, we recommend that the Otitis Media with Effusion guideline be withdrawn.

There have been several new clinical trials of preventive strategies and therapies for acute low back pain that have been published since the release of the guideline. Specifically, there are new clinical trials reporting no or only modest efficacy of educational back schools at the work site or the use of lumbar corsets in preventing acute low back pain, and of epidural steroid injections as symptomatic therapy. These data would likely further temper some of the modest recommendations for efficacy made in the guideline. All three of these topics were communicated with lukewarm enthusiasm in the guideline (back schools "may be effective", lumbar corsets "may reduce" time lost from work, and epidural steroids are an "option" for treatment) and all received grade C recommendations. As we judged the majority of the guideline to be still valid, and as the new information involves aspects of the guideline that we judge to be of lesser importance in producing good patient outco! mes than other valid statements in the areas regarding diagnosis, treatment and education, we recommend that the Acute Low Back Problems guideline be retained, with the new evidence appended.

All of the statements in the Post-Stroke Rehabilitation guideline were judged to be still valid. However, there has been considerable new evidence further supporting the value of stroke units in improving outcomes, including mortality, among stroke patients. The guideline statement concerning stroke units state: "Whenever possible, patients with acute strokes should receive coordinated diagnostic, acute management, preventive, and rehabilitative services." This recommendation received a grade A level of evidence and a rating of expert opinion equal to "consensus", which was not the strongest level of support (that designation being "strong consensus"). The new evidence would likely increase the level of support for the recommendation for stroke units, possibly to a point consistent with a recent recommendation from the Brain Attack Coalition stating that centers that "intend to provide care beyond the hyperacute period... should provide such care in a stroke unit setting"! ¹³ We judged that this information should be made known to clinicians, and we recommend the Post-Stroke Rehabilitation guideline be retained with the new evidence appended.

The Recognition and Initial Assessment of Alzheimer's Disease and Related Dementias guideline was limited in scope because of a "lack of unequivocally effective treatment". Since the guideline was released there have been randomized trials supporting modest efficacy for cholinesterase inhibitors in patients with Alzheimer's Disease. Based on this evidence, some newer guidelines recommend the use of cholinesterase inhibitors in select patients. It is arguable whether these data represent "unequivocal evidence" of efficacy, however we judged that clinicians should be aware of this new information. We recommend that the Alzheimer's guideline be retained, with this new evidence appended.

We acknowledge that our recommendations for these nine guidelines involve a substantial amount of judgement on our part. Others, faced with the same evidence, may reach different conclusions. In this report, we have made explicit the evidence and rationale for our decisions, so that the readers can judge for themselves the soundness of our conclusions.