Evidence-based guidelines: application to clinical practice. (Introduction to Special Issue).

**********

This special issue of Physical Therapy contains 4 clinical practice guidelines for physical therapist management of musculoskeletal conditions. These clinical practice guidelines were developed using systematic reviews of the literature and provide the reader with evidence suitable for evidence-based physical therapist practice. It is the responsibility of the individual clinician, however, to determine the applicability of the evidence from these guidelines to individual patients. The purposes of this article are to review the benefits and limitations of clinical practice guidelines in general and to discuss strategies for application to individual patients. Although users of clinical practice guidelines may include clinicians, payers, regulatory agencies, and patients, the focus of this article is on uses by physical therapy clinicians.

WHAT ARE CLINICAL PRACTICE GUIDELINES?

Clinical practice guidelines are defined by the Institute of Medicine as systematically developed statements to assist practitioner and patient decisions about appropriate health care for specific clinical circumstances. (1) In the past 10 years, the number of practice guidelines appearing in the medical literature has increased rapidly. For example, of the nearly 5,500 articles indexed by MEDLINE through 2000 according to the publication type "practice guideline," over 5,000 were published between 1991 and 2000, with almost 700 of these articles appearing in 2000.

One reason often given for the development of practice guidelines is to improve the quality of patient care. (2) The consolidation and gauging of recent evidence from the literature about the relative effectiveness of various treatment strategies should result in clinicians changing their attitudes about treatment choices, which will lead to behavior changes (eg, greater use of treatments with known effectiveness) and ultimately improved patient outcomes. Guidelines attempt to accomplish this by making summary recommendations based on available evidence.

The appearance of terms such as "systematic" within guidelines does not ensure that the guidelines were developed using a structured review format. Methods to review the literature vary. Therefore, the minimal criteria for any practice guideline should include explicit statements that explain the process of creating the recommendations and the systematic search and gauging of the evidence. Absence of a description of the process of development should cause users to question the validity of the guideline.

Based on the current literature and their experience in guideline development, Shekelle et al (3) delineate 5 steps in evidence-based guideline development. First, the subject area and the need for the guideline are identified and defined. Failure to identify and define the subject area can lead to a guideline that does not specifically address the condition or procedure targeted. Second, a development group is created and the members of the group and their specific roles are identified. Ideally, the group should contain representatives of all identified stakeholders whose activities would be affected by the guideline. Third, rules for identifying and assessing evidence are determined before searching for evidence. The methods for identifying and assessing evidence should ensure that all potential sources of evidence that relate to the defined subject area are included. Fourth, a method for translating evidence from different types of studies and from studies of different methodological quality into a recommendation needs to be specified. In addition to considering the evidence, the recommendations need to consider benefits of the intervention, limitations associated with the intervention, the population for whom the recommendation is most applicable, costs, and factors related to the health care system. The fifth and final step in guideline development is to specify a method of review and process to update the guideline. A guideline can only be as good as the evidence on which it is based. As new evidence is identified, the recommendations may no longer be appropriate. Therefore, revisions to the guideline are necessary to ensure that the guideline is based on the best current evidence. (2)

Evidence-based clinical practice guidelines are 1 of 4 types of guidelines that have been described in the literature. (4,5) The simplest type of guideline is based on expert consensus. The development of expert-based guidelines is less time-consuming and less costly than that of other types of guidelines. Since these guidelines are based on the opinion of experts, and not on evidence, a lower level of applicability is usually associated with them. A limitation of expert-based guidelines is that they may reflect only the opinions of the developers and their profession-specific biases.

Another type of clinical practice guidelines is outcome based. (4,5) These guidelines include a measure of the effectiveness of evidence-based recommendations within the guideline to determine whether the recommendation improved the quality of care. Various methods, such as meta-analysis, decision analysis, or cost-effectiveness analysis, may be used to determine the effectiveness of the recommendation. Assessing the benefits adds time to the development process, but allows for a greater applicability of the recommendations of this type of guideline. Measurement of the effectiveness provides additional evidence for the guideline and may identify any needs for revision. (4,5)

The most rare type of clinical practice guidelines is preference based. (4,5) The methods of evidence-based and outcome-based guidelines are combined with patient preferences for possible outcomes of the interventions. Patient preferences have been described as "a person's views about quality of life experienced with different health states" and can be used in applying treatment decisions to individual patients. (4-6) The inclusion of patient preferences makes these guidelines very difficult to create, but provides the only form of guidelines that takes into account the variability in the values of individual patients and includes the patients' values in the decision-making process.

Very few clinical practice guidelines have included patient preferences. One guideline that has done so is the Agency for Health Care Policy and Research's guideline for the management of benign prostatic hyperplasia. (7) Benign prostatic hyperplasia affects the quality of life more than the quantity, and treatment of benign prostatic hyperplasia is associated with risk. The benign prostatic hyperplasia practice guideline included the preferences of men with mild, moderate, or severe benign prostatic hyperplasia and men without signs of prostatic disease to assess the various health states associated with the treatment recommendations. Although limited in scope by a small sample size, the preferences obtained in this guideline from men with and without prostatic disease were variable. The variations were likely due to individual preferences for different treatments, the degree of perceived bothersomeness of the symptoms, and how each individual views the risks associated with therapy. This guideline concluded that patients' choices need to be elicited and that patients and health care providers should share the decision making regarding the treatment strategy for benign prostatic hyperplasia.

The categorization of the 4 types of practice guidelines refers to the inclusion of different types of information in the development of the guideline. Outcomes for different interventions and patient preferences for different intervention options are still important for the use of evidence-based guidelines, even though outcomes and preferences may not be included within that type of guideline. Assessment of the effectiveness of all guidelines is crucial in determining whether the recommendation of the guideline has improved the quality of patient care. Data obtained from measuring the implementation of the recommendation provide additional evidence for the guideline that needs to be considered in the revision of the guideline. This additional evidence may be complementary or contradictory and may identify areas where additional evidence is needed. Patient preferences for the intervention choices of evidence-based guidelines are important to consider in clinical decision making.

SYSTEMATIC REVIEWS

Recommendations of practice guidelines typically are enhanced by the inclusion of a systematic review of the current literature. A systematic review is a summary of the medical literature using explicit methods to guide systematic searches, critical appraisal, and synthesis of the literature on a specific issue. (8) The systematic review may combine the studies it compares quantitatively to create a meta-analysis, or the comparison of different studies may be qualitative. (9) Systematic reviews of evidence can be a benefit to busy clinicians as evidence can be combined and summarized in one source. Systematic reviews and meta-analyses may be included in a clinical practice guideline, but are not required and have distinct characteristics from practice guidelines. One critical distinction between clinical practice guidelines, systematic reviews, and meta-analyses is that a guideline makes a recommendation based on all the issues that may influence a clinical decision in addition to the identification of the available evidence. (10,11)

One strength of a systematic review is that the methods for its development are explicit. Each step in the process is specifically described, including the strategy to search for evidence, the criteria including and excluding studies in the review, and the information that is obtained from each study. Two advantages of describing and following explicit methods of a systematic review, analogous to describing and following explicit methods in an experimental study, are (1) reproducibility of the methods is improved and (2) there are greater controls for investigator bias.

A number of misconceptions about the relationship of systematic reviews to traditional narrative reviews may exist, such as:

* systematic reviews are the same as narrative reviews, only larger and more comprehensive;

* systematic reviews include only randomized trials;

* systematic reviews require the use of meta-analysis;

* systematic reviews are a substitute for doing additional good-quality studies. (12)

First, systematic reviews may actually include fewer references than narrative reviews due to the application of stringent inclusion criteria. Although more databases may be searched, fewer studies may meet the specified quality requirements. Next, systematic reviews may include studies other than randomized controlled trials, such as cohort studies and case-control studies. These reviews, however, should describe how these other types of studies are identified and should present criteria for how the results of these other types of studies will be reviewed.

A systematic review does not have to include the use of meta-analysis. The use of meta-analysis in systematic reviews is feasible only when the results can be combined quantitatively. When the results cannot be combined--including when studies of different designs are compared, when studies have used different methods, and when different outcomes are measured--meta-analysis should not be used, and the studies should be compared qualitatively.

Systematic reviews may be misconstrued as eliminating the need for further experimental studies by providing definitive evidence for clinical practice. In reality, systematic reviews attempt to summarize the current evidence, which rarely provides a definitive answer. Providing a summary of current evidence can help to identify needs and direct future research by specifically identifying where research is currently lacking or less than optimal. Other misconceptions of systematic reviews are discussed in a recent article by Petticrew. (12)

VARIABILITY IN CLINICAL PRACTICE GUIDELINES AND SYSTEMATIC REVIEWS

As the popularity of practice guidelines and systematic reviews increases, the potential for discrepancies may exist when multiple practice guidelines are published on the same topic. On some topics, such as spinal manipulation for low back problems, the number of systematic reviews that have been published is greater than the number of studies that have been reviewed. (13) Of the 51 systematic reviews published up to 1993 on spinal manipulation, the summary of evidence was in favor of manipulation in 34 reviews and the evidence was neutral to manipulation in 17 reviews. With an increasing number of guidelines and conflicting messages, a clinician may be confused regarding what recommendations to follow, and the applicability of the recommendations may be lost.

In deciding among competing practice guidelines, the first step is to determine the validity of the guidelines and the systematic review of evidence on which it is based. (11,14) A valid guideline contains statements that explicitly describe how the recommendations of the guideline were developed and how the systematic review of evidence was conducted. The presence of these statements, however, does not ensure validity. The user should determine whether the statements were followed by the authors of the guideline. A practice guideline that deviated from its stated methods may be considered less valid as compared with a practice guideline that followed its stated methods.

Recent evidence suggests that the quality of clinical practice guidelines has improved over time, but that many guidelines and systematic reviews have not adhered to established methodological standards. (15,16) Examples of some of the most frequently violated standards include not specifying a method for identifying evidence, not describing methods of combining evidence, not discussing the role of value judgments in making recommendations, and not specifying a date for scheduled review of the guideline. The developers of guidelines should provide information within the guideline that allows the user to determine whether the guideline is valid without having to read the individual studies used in its creation. A valid guideline should consider all of the important treatment choices and consequences, provide publication dates for evidence considered and for the creation of the final recommendations, and undergo peer review, in addition to providing an explicit description of development of the guideline, making of recommendations, and identification of evidence. (11)

A particular recommendation of a guideline can only be appreciated in the context of all available and alternative interventions. A guideline that does not provide a thorough description of this context, therefore, will not benefit the user in understanding the recommendation. By providing dates for the evidence and recommendations contained in a guideline, the user can determine whether evidence published after these dates needs to be considered along with the guideline. Peer review can improve the validity of a guideline by allowing people external to the guideline developers to judge whether the recommendations are reasonable and are applicable for practice. Clinicians should seek and prefer guidelines that meet these various criteria for validity over guidelines that do not meet these criteria. Caution is required in the use of guidelines with weaker validity because they may fail to provide information necessary to improve the quality of care.

After determining the validity of a practice guideline, multiple valid guidelines with different recommendations may exist for a clinical problem. Differing recommendations may result from differences in the methods used to make the recommendations or to systematically review the literature, or from a combination of both. To understand the differences between multiple practice guidelines, the user must identify whether this difference is in the identification of the evidence, the creation of the recommendation, or a combination of factors related to both the evidence and the recommendations. (17) In an attempt to standardize methods to create recommendations, the Scottish Intercollegiate Guidelines Network (SIGN) has recently proposed a system for grading recommendations that allows the available evidence to be linked to the recommendations that take into account evidence from different types of studies (ie, not only randomized controlled trials) and evidence from studies of different methodological quality. (18) Although promising, the methods of the SIGN grading system have yet to be validated.

Methods used to develop recommendations may differ according to the stakeholders involved in the guideline development, how expert opinion is used, and how the recommendations are expressed. To develop recommendations, evidence often is combined with value judgments of the developers. Understanding the perspective of the developers, regardless of whether they are using the patient's perspective, a societal perspective, or their own perspective, may result in different recommendations being made from the same body of evidence.

In some cases, opposing recommendations are made when no supportive evidence or minimal evidence from a systematic review exists. For example, one guideline may label an intervention as "not recommended" based on the absence of evidence, whereas another guideline may label an intervention as "optional" or may be unable to make a recommendation based on the lack of evidence. Users of guidelines should keep in mind that a lack of evidence is not the same thing as evidence against an intervention. Ultimately, it is the responsibility of the consumer to determine the implications of these differences for application to patient care.

Different recommendations may also result in practice guidelines when different methods were used to systematically review the literature. Purposes for the creation of guidelines may differ, therefore resulting in different bodies of evidence being searched for the systematic review, which could lead to different conclusions. Other areas of difference may be the inclusion and exclusion criteria for the systematic review, the quality of the trials included in the review, and whether one review includes more recent evidence. The reader is referred to an algorithm published by Jadad et al (17) (Figure) to use when deciding between systematic reviews that disagree. Using a tool to help interpret discordant systematic reviews may make it easier for clinicians to chose which of the reviews, if any, is more applicable to their patients.

[FIGURE OMITTED]

IMPLEMENTATION OF GUIDELINES

The development of clinical practice guidelines that summarize the evidence and provide recommendations is one step in fostering evidence-based practice. One of the suggested benefits of practice guidelines is to improve outcomes by providing a better quality of care. The presence of guidelines does not guarantee that the recommendations of the guidelines will improve the outcome for an individual patient, however. For example, the Agency for Health Care Policy and Research introduced Acute Low Back Problems in Adults (Clinical Practice Guideline No. 14) in 1994. (19) Studies from various disciplines, including general practice, (20) radiology, (21,22) and chiropractic, (23) however, have determined that this guideline has not been fully implemented.

The implementation of guidelines requires that more be done than simply the publication of guidelines. A systematic review of the effectiveness of printed educational materials in changing the behavior of health care professionals suggests that publication of clinical practice guidelines is rarely effective by itself. (24)

Feder et al (25) suggest the use of multifaceted methods to disseminate and implement guidelines, as no single strategy has been found to be effective in ensuring that guidelines are used in clinical practice. Part of an implementation strategy may include identification of barriers to the use of guidelines. Barriers to implementation may be environmental, financial, cultural, or related to a lack of knowledge regarding performance. For example, clinicians may not be aware that they are relying on beliefs instead of evidence to deliver physical therapy, which in turn results in their delivering less-than-optimal care. Feedback given to the clinicians on their performance and outcomes may help them to identify areas of practice that require updated knowledge, which may result in them seeking a clinical practice guideline.

Practice guidelines themselves may be another barrier to their implementation. Guidelines that are confusing or contradictory with current practice may meet greater resistance in implementation than guidelines that are consistent with current practice. Grol et al (26) found that guidelines with clear recommendations, noncontroversial recommendations, and the use of evidence-based recommendations were more likely to be followed than guidelines with recommendations that were not clear, were controversial, or were based on opinion.

APPLICATION TO INDIVIDUAL PATIENTS

Most evidence included in practice guidelines is from experimental studies of groups of patients. In contrast, clinicians are required to make decisions for individual patients. Some clinicians, therefore, may immediately denounce the recommendations of any clinical practice guideline on the basis that their patient is different than the patients in the studies. Instead, a starting point for the application of any evidence to an individual patient should be determining how the patient is similar to the patients in the study groups and not how the patient is different. (7) A practice guideline, therefore, should provide a thorough and accurate description of the population on which a recommendation is based to allow a clinician to identify similarities. The description within the guideline should provide enough information to minimize the need for the user to return to the original studies to find this information. Rarely will a patient have all of the characteristics of the group of study patients. By focusing on the similarities, however, a clinician may begin to relate how the evidence included in a practice guideline may apply to an individual patient. The use of evidence in practice is to help reduce uncertainty in decision making, not to eliminate it. Elimination of all uncertainty is idealistic and may be impossible to achieve.

The clinician should identify a practice guideline that is most similar to the clinical situation of the patient. In order to do this, however, the guideline needs to contain information that clearly defines the clinical situation for which the recommendations are applicable. For example, a clinician in a primary care setting managing a client's acute low back pain should more likely follow a recommendation developed from evidence that was obtained from people with low back pain seen in acute care settings than a recommendation developed from evidence that was obtained from people with chronic low back pain seen in specialized care centers.

In addition to comparing the similarity of a patient with the research subjects, the feasibility of implementing the intervention in the guideline should be determined. Perhaps environmental and financial barriers exist that may limit the implementation of an exercise approach that requires additional space and equipment. The implementation of an intervention recommended by a guideline also may be limited by the clinician's need to obtain additional training to master the skills required for a technique in which the clinician may not be proficient, such as manipulation for low back problems.

To apply evidence to an individual patient, the expected outcome of treatment should be similar to the outcome that is valued by the patient. In order to accomplish this, the clinician not only should know the evidence regarding different expected outcomes but should elicit the value that the patient ascribes to different outcomes. To elicit the patient's value for different outcomes, the clinician first needs to fully inform the patient about these outcomes. Unfortunately, recent evidence suggests that some physical therapists have not taken full advantage of the potential for patient participation during treatment planning. (27)

Formal quantitative methods, such as decision analysis and the likelihood of help versus harm, also can be used in applying evidence to individual patients. (8,9) A decision analysis is the application of explicit methods that quantify and combine prognoses, treatment effects, and patient values to analyze a decision under conditions of uncertainty. (9) The likelihood of help versus harm combines the number-needed-to-treat statistic, (28) the clinician's perspective of the patient's risk, and the patient's perception of the severity of the condition being prevented and any potential side effects. (8) Application of one of these quantitative tools may be appropriate when all of the values for the calculations are known and when the resulting calculation will influence the decision made.

SUMMARY

According to Sackett et al, (8) the definition of evidence-based medicine is the integration of best research evidence with clinical expertise and patient values. The clinical practice guidelines in this issue of Physical Therapy identify evidence for the physical therapist management of certain musculoskeletal conditions. The responsibility remains with the clinician to combine this evidence with clinical expertise and patient values in managing individual patients and achieving optimal outcomes.

References

(1) Field MJ, Lohr KN. Clinical Practice Guidelines: Directions of a New Program. Washington, DC: National Academy Press, 1990.

(2) Woolf SH, Grol R, Hutchinson A, et al. Clinical guidelines: potential benefits, limitations, and harms of clinical guidelines. BMJ. 1999;318: 527-530.

(3) Shekelle PG, Woolf SH, Eccles M, Grimshaw J. Clinical guidelines: developing guidelines. BMJ. 1999;318:593-596.

(4) Eddy DM. Practice policies: where do they come from? JAMA. 1990;263:1265, 1269, 1272.

(5) Owens DK. Spine update: patient preferences and the development of practice guidelines. Spine. 1998;23:1073-1079. 6

(6) Froberg DG, Kane RL. Methodology for measuring health-state preferences, II: scaling methods. J Clin Epidemiol. 1989;42:459-471.

(7) McConnell JD, Barry MJ, Bruskewitz RC. Benign Prostatic Hyperplasia: Diagnosis and Treatment. Clinical Practice Guideline No. 8. Rockville, Md: Agency for Health Care Policy and Research, Public Health Service, US Department of Health and Human Services; 1994. AHCPR publication no. 94-0582.

(8) Sackett DL, Strauss SE, Richardson WS, et al. Evidence-Based Medicine: How to Practice and Teach EBM. 2nd ed. Edinburgh, Scotland: Churchill Livingstone; 2000.

(9) Petitti DB. Meta-Analysis, Decision Analysis, and Cost-Effectiveness Analysis: Methods for Quantitative Synthesis in Medicine. New York, NY: Oxford University Press; 2000.

(10) Cook DJ, Greengold NL, Ellrodt AG, Weingarten SR. The relation between systematic reviews and practice guidelines. Ann Intern Med. 1997;127:210-216.

(11) Hayward RS, Wilson MC, Tunis SR, et al. Users' guides to the medical literature, VIII: how to use clinical practice guidelines, A: are the recommendations valid? The Evidence-Based Medicine Working Group. JAMA. 1995;274:570-574.

(12) Petticrew M. Systematic reviews from astronomy to zoology: myths and misconceptions. BMJ. 2001;322:98-101.

(13) Assendelft WJ, Koes BW, Knipschild PG, Bouter LM. The relationship between methodological quality and conclusions in reviews of spinal manipulation. JAMA. 1995;274:1942-1948.

(14) Oxman AD, Cook DJ, Guyatt CH. Users' guides to the medical literature, VI: how to use an overview. JAMA. 1994;272:1367-1371.

(15) Shaneyfelt TM, Mayo-Smith MF, Rothwangl J. Are guidelines following guidelines? The methodological quality of Clinical practice guidelines in the peer-reviewed medical literature. JAMA. 1999;281: 1900-1905.

(16) Jadad AR, Cook DJ, Jones A, et al. Methodology and reports of systematic reviews and meta-analyses: a comparison of Cochrane reviews with articles published in paper-based journals. JAMA. 1998; 280:278-280.

(17) Jadad AR, Cook DJ, Browman GP. A guide to interpreting discordant systematic reviews. CMAJ. 1997;156:1411-1416.

(18) Harbour R, Miller J. A new system for grading recommendations in evidence based guidelines. BMJ. 2001;323:334-336.

(19) Bigos S, Bowyer O, Braen G, et al. Acute Low Back Problems in Adults. Clinical Practice Guideline No. 14. Rockville, Md: Agency for Health Care Policy and Research, Public Health Service, US Department of Health and Human Services; 1994. AHCPR publication no. 95-0642.

(20) Little P, Smith L, Cantrell T, et al. General practitioners' management of acute back pain: a survey of reported practice compared with clinical guidelines. BMJ. 1996;312:485-488.

(21) Freeborn DK, Shye D, Mullooly JP, et al. Primary care physicians' use of lumbar spine imaging tests: effects of guidelines and practice pattern feedback. J Gen Intern Med. 1997;12:619-625.

(22) Suarez-Almazor ME, Belseck E, Russell AS, Mackel JV. Use of lumbar radiographs for the early diagnosis of low back pain: proposed guidelines would increase utilization. JAMA. 1997;277:1782-1786.

(23) Shekelle PG, Coulter I, Hurwitz EL, et al. Congruence between decisions to initiate chiropractic spinal manipulation for low back pain and appropriateness criteria in North America. Ann Intern Med. 1998; 129:9-17.

(24) Oxman AD, Thomson MA, Davis DA, Haynes RB. No magic bullets: a systematic review of 102 trials of interventions to improve professional practice. CMAJ. 1995;153:1423-1431.

(25) Feder G, Eccles M, Grol R, et al. Clinical guidelines: using clinical guidelines. BMJ. 1999;318:728-730.

(26) Grol R, Dalhuijsen J, Thomas S, et al. Attributes of clinical guidelines that influence use of guidelines in general practice: observational study. BMJ. 1998;317:858-861.

(27) Baker SM, Marshak HH, Rice CT, Zimmerman GJ. Patient participation in physical therapy goal setting. Phys Ther. 2001;81:1118-1126.

(28) Dalton GW, Keating JL. Number needed to treat: a statistic relevant for physical therapists. Phys Ther. 2000;80:1214-1219.

DA Scalzitti, PT, MS, OCS, is Clinical Instructor, Department of Physical Therapy, University of Illinois at Chicago, 1919 W Taylor St (M/C 898), Chicago, IL 60612 (USA) ([e-mail]), and Specialist in Physical Therapy, Department of Physical Therapy, University of Illinois Hospital, Chicago, IL 60612. Address all correspondence to Mr Scalzitti at the first address.