- Email: [email protected]
Factors influencing hospital implementation of acute pain management practice guidelines6
Original Contributions Factors Influencing Hospital Implementation of Acute Pain Management Practice Guidelines H. Joanna Jiang, PhD,* Robert S. Lagasse, MD,† Kathleen Ciccone, RN, MBA,‡ Michael S. Jakubowski, MD,§ Eric M. Kitain, MD储 Healthcare Association of New York State, Albany, NY
*Senior Researcher, Healthcare Association of New York State, Albany, NY. Currently Social Scientist, Agency for Healthcare Research and Quality, Rockville, MD. †Associate Professor of Clinical Anesthesiology, Montefiore Medical Center, Bronx, NY ‡Vice President, Division of Quality & Research Initiatives, Healthcare Association of New York State, Albany, NY §Vice President, Medical Affairs, Ellis Hospital, Schenectady, NY 储Chairman of Anesthesiology, Long Island Jewish Medical Center, New Hyde Park, NY Address correspondence and reprint requests to Dr. Lagasse at the Department of Anesthesiology, Montefiore Medical Center – Weiler Division, 1825 Eastchester Rd., Bronx, NY 10461. E-mail: [email protected] This study was part of a larger project that was funded by the New York State Department of Health, Albany, NY, as recommended by the Task Force on Clinical Guidelines and Technology Assessment (Grant #TF050). Received for publication October 4, 2000; revised manuscript accepted for publication March 12, 2001.
Study Objective: To identify factors that may influence the implementation of acute pain management guidelines in hospital settings. Design: Two questionnaire surveys. Setting: Healthcare Association of New York State, Albany, NY. Measurement: The surveys were administered to 220 hospitals in New York State regarding their acute pain management practices and resources available. One survey was addressed to each hospital’s chief executive officer (CEO) and the second survey was addressed to the clinical director of the Department of Anesthesiology or Acute Pain Service. The barriers and incentives to guideline implementation identified by CEOs were analyzed using factor analysis. Logistic regression was employed to determine predictors of guideline implementation by linking the CEOs’ survey data with the clinical directors’ report of guideline usage. Main Results: According to clinical directors, only 27% of the responding hospitals were using a published pain management practice guideline. Factors predictive of guideline implementation include resource availability and belief in the benefits of using guidelines to improve quality of care or to achieve economic/legal advantages. Guideline implementation, however, does not necessarily include applying all key elements recommended by the federal Agency for Healthcare Research and Quality (formerly Agency for Health Care Policy and Research) guideline. For example, a collaborative, interdisciplinary approach to pain control was used in only 42% of the hospitals, and underutilization of nonpharmacologic therapies to control pain was widespread. Resource availability, particularly staff with expertise in pain management and existence of a formal quality assurance program to monitor pain management, was significantly predictive of compliance with key guideline elements. Conclusions: Resource availability significantly influences the implementation of pain management practice guidelines in hospital settings. Implementation is often incomplete because various factors affect the feasibility of individual guideline elements and may explain the varying results that guidelines have had on clinical practices. © 2001 by Elsevier Science Inc. Keywords: Acute pain management, clinical practice guidelines, quality improvement.
Journal of Clinical Anesthesia 13:268 –276, 2001 © 2001 Elsevier Science Inc. All rights reserved. 655 Avenue of the Americas, New York, NY 10010
0952-8180/01/$–see front matter PII S0952-8180(01)00268-9
Acute pain management guidelines: Jiang et al.
Introduction Pain is described as “an unpleasant sensory and emotional experience associated with actual or potential tissue damage.”1 Inadequate treatment of postoperative and other acute pain is well documented in the medical literature.2–5 In response, several national entities, such as the federal Agency for Healthcare Research and Quality (AHRQ; formerly Agency for Health Care Policy and Research, AHCPR), the American Pain Society (APS), and the American Society of Anesthesiologists (ASA), have developed pain management practice guidelines to facilitate better assessment and treatment of pain.6 – 8 Although implementation of these practice guidelines was found to have a favorable impact on improving patient outcomes and satisfaction,9 –12 the guidelines have not been adopted extensively at the local level. The importance of addressing pain management for improving quality of care is well reflected by the current initiative of the Joint Commission on Accreditation of Healthcare Organizations (JCAHO), which will use guideline-based measures of pain management in its survey process in the year 2001. This fact presents strong incentives for hospitals and other health care organizations to adopt pain management practice guidelines. Inevitably, there will be increasing interest in what influences the implementation process and how to ensure the successful adoption of guidelines. A vast amount of research has been conducted to examine factors that affect the implementation of clinical practice guidelines.13 Variables that have been identified as important to guideline implementation include: 1) attributes of guidelines14; 2) physician characteristics, including attitudes about guidelines15–17; specialty training18; and gender19; 3) practice setting15,20; 4) patient factors21; and 5) incentives related to financial, legal, and regulatory issues. Nevertheless, to date, few empirical studies have addressed barriers to the adoption of pain management practice guidelines. Most guidelines give little attention to the feasibility of implementing the recommendations across a wide variety of health care settings. In reviewing the AHRQ/AHCPR acute pain management guideline, Carr and colleagues suggested that the following local resource shortfalls might limit implementation22: 1) lack of staff trained and knowledgeable in pain management, 2) lack of specially trained nurses or psychologists to deliver nondrug interventions, 3) lack of formal quality assurance or other institutional programs to monitor pain management, 4) absence of equipment necessary to deliver some of the interventions, 5) lack of specialists with the requisite expertise to insert and monitor epidural catheters, and 6) lack of resources to provide the education to clinicians to implement the guideline. Prior studies on clinical practice guidelines tend to focus on physicians’ perspectives on guideline implementation, while little is known about health executives’ perspectives on barriers to guideline adoption. Good understanding and support from hospital administration is imperative to the successful implementation of acute pain management practice guidelines. This study seeks to
facilitate understanding of select issues related to implementation of a clinical practice guideline on acute pain management. Three questions were addressed through surveys of hospital chief executive officers (CEOs) and clinical directors. First, what are the hospital CEOs’ perceptions of factors influencing pain management guideline implementation? Second, how do current pain management practices, as reported by the clinical directors, compare to what is recommended by the guideline? Third, do hospital CEOs’ perceptions of issues related to guideline implementation predict the adoption of guideline elements as reported by the clinical directors?
Materials and Methods To identify factors that may influence the decision by a hospital to implement acute pain management practice guidelines, two sets of surveys were developed, pilot tested, and mailed to 220 hospitals in New York State. One survey was addressed to each hospital’s CEO and the second survey was addressed to the clinical director of the Departments of Anesthesiology or Acute Pain Service. The CEOs were asked the extent to which they agree or disagree with statements about the accessibility of staff trained and knowledgeable in all modalities of pain management, availability of equipment necessary to deliver the recommended interventions, reimbursement issues, and possible benefits to the hospital of implementing a pain management guideline. The clinical directors were asked about the type of acute postoperative pain management program being instituted, the use of a nationally published clinical practice guideline, and current pain management practices. Questions on current pain management practices assessed compliance with elements of the AHRQ/ AHCPR acute pain management practice guideline. Responses on both surveys were graded on a five-point Likert scale ranging from “strongly agree” to “strongly disagree.” Responses to the CEO survey were summarized using factor analysis to identify underlying dimensions (called “factors”) that were measured by the 12 survey questions. Factor analysis is a multivariate statistical technique that analyzes the interrelationship among a large number of variables and summarizes these variables in terms of their common underlying factors.23 It allows one to generate a more parsimonious description of the survey results. Factor analysis has been widely used in health services research for validating survey instruments. For example, Ware and colleagues at the New England Medical Center used factor analysis to validate the eight physical and mental health components measured by the SF-36 Health Survey.24 Zaslavsky and colleagues at Harvard Medical School used factor analysis to summarize the 39-item Consumer Assessment of Health Plans Survey (CAHPS) into four subjects, including interactions with clinical staff, customer service, access to care, and advice on healthpromoting activities.25 The second purpose of performing factor analysis in this study was to reduce the original 12 survey questions to a smaller number of factors that subsequently would be used in the logistic regression J. Clin. Anesth., vol. 13, June 2001
269
Original Contributions
Table 1. Descriptive Statistics for Hospital Characteristics
Hospital Characteristics Bed size ⬍ 250 beds 250–499 beds 500 ⫹ beds Teaching status Major teaching Minor teaching Nonteaching Location Urban Rural Sponsorship Voluntary Public Proprietary
Clinical CEO Director Survey Survey Respondents Respondents All New York (nⴝ137) (nⴝ135) State Hospitals
49.6% 32.8% 17.5%
49.2% 34.1% 16.7%
55.8% 30.4% 13.8%
27.0% 24.8% 48.2%
25.8% 25.0% 49.2%
23.1% 26.9% 50.0%
83.9% 16.1%
83.3% 16.7%
84.2% 15.8%
90.5% 9.5% –
90.2% 9.8% –
85.8% 10.8% 3.5%
CEO ⫽ chief executive officer.
analysis. Each factor represents a composite of all the variables related to that particular factor. Logistic regression analysis was conducted to validate the significance of hospital CEOs’ perceptions of factors related to guideline implementation by using their responses to predict the usage of pain management guidelines as reported by the clinical directors. The outcome variable was set equal to 1 if the clinical director reported using a guideline, and equal to 0 otherwise. The model is: log[P/(1 ⫺ P)] ⫽ 0 ⫹ 1X ⫹ 2Y, where P is the probability of implementing a guideline, and X is a vector of hypothetical factors derived from the factor analysis of the CEO survey, while Y is a vector of hospital characteristics including bed size, teaching status, urban/rural location, and sponsorship. The results of the logistic regression were used to identify factors significantly related to the implementation of a pain management practice guideline. Similarly, logistic regression analysis was used to further identify those factors, as perceived by hospital CEOs, that significantly predict the likelihood of complying with key elements of the AHRQ/AHCPR guideline.
Results The results are presented in three subsections. The first subsection describes the results of the CEO survey and the clinical director survey. The second subsection reports how the results of the 12 questions on the CEO survey were summarized into four areas of perceived barriers/ incentives to guideline implementation: a) availability of institutional resources, b) economic/legal incentives for guideline implementation, c) potential impact on patient outcomes, and d) concerns on reimbursement. The last subsection describes how a hospital CEO’s perception 270
J. Clin. Anesth., vol. 13, June 2001
of issues on guideline implementation was related to the actual adoption of pain management practice guidelines in the hospital and compliance of key guideline elements.
The CEO and the Clinical Director Surveys The surveys were administered to 220 hospitals in New York State in 1996 and 1998 as part of a three-year study funded by the New York State Department of Health to evaluate the impact of pain management practice guideline implementation on clinical practices and patient outcomes. One hundred thirty-seven (62.3% response rate) of the CEOs and 135 (61.4%) of the clinical directors responded to the first round of surveys in 1996 while only 90 hospitals responded to the second round of surveys in 1998. No significant difference was found between the results of the second round of surveys and those of the initial surveys. Therefore, we chose to present the results of the first round of surveys which had a larger hospital sample size. Hospital characteristics are summarized in Table 1. There was no statistically significant difference in any of the characteristics between the survey sample and the New York State hospital population, suggesting that the survey sample well represented New York hospitals. The results of the CEO survey are presented in Table 2. One hundred three (74%) of the CEOs responding agreed that their staff were trained and knowledgeable in pain management, but only 47 (34%) indicated that they had enough specially trained staff to deliver nondrug interventions. One hundred twenty-three (90%) CEOs also agreed that they had the necessary equipment. A majority of the responding hospitals (69%) had formal quality assurance (QA) programs to monitor pain management. The CEOs agreed that implementation would improve market value to payers (77%), risk of medical liability may be reduced for practitioners following the guideline (67%), and that patients would have fewer complications, and in some cases, shorter hospital stays (80%). Eighty-nine (65%) CEOs, however, agreed that reimbursement was an issue to consider before implementing pain management practice guidelines. Of the clinical directors responding to the survey (Table 3), 76 (56%) stated that they had a comprehensive acute pain management program, but only 37 (27%) used a published clinical practice guideline to establish their program. A collaborative, interdisciplinary approach to pain control was used for all or most inpatients in 42% of the hospitals responding. Individualized proactive pain control plans were developed preoperatively in all or most inpatients in 39% of the hospitals. Close to 50% of the hospitals reported using a formal postoperative pain assessment tool. Although 89% to 100% of the hospitals used pharmacologic interventions such as opioids, local anesthetics, and nonsteroidal antiinflammatory drugs (NSAIDs), the percentage of hospitals offering nonpharmacologic interventions such as patient education (69%), transcutaneous electrical nerve stimulation (TENS) (52%), and relaxation therapy (36%), was considerably lower.
Acute pain management guidelines: Jiang et al.
Table 2. Results of the Hospital CEO Survey (n ⫽ 137) 1. Our hospital staff (including physicians, nurses, psychologists, and others) is generally trained and knowledgeable in pain management. 18.4% 55.1% 5.9% 19.9% 0.7% Strongly Mostly Unsure Mostly Strongly Agree Agree Disagree Disagree 2. Our hospital has the equipment necessary to deliver some of the interventions for pain relief (including various drugs and pumps for patient-controlled analgesia.) 59.1% 31.4% 0.7% 7.3% 1.5% Strongly Mostly Unsure Mostly Strongly Agree Agree Disagree Disagree 3. Our hospital has enough specialists with the requisite expertise to insert and monitor epidural catheters. 57.4% 28.7% 1.5% 9.6% Strongly Mostly Unsure Mostly Agree Agree Disagree
2.9% Strongly Disagree
4. Our hospital has enough specially trained nurses and psychologists to deliver nondrug interventions. 8.1% 25.7% 30.1% 26.5% Strongly Mostly Unsure Mostly Agree Agree Disagree
9.6% Strongly Disagree
5. Our hospital has a formal quality assurance or other institutional program to monitor pain management. 29.2% 39.4% 10.9% 16.1% Strongly Mostly Unsure Mostly Agree Agree Disagree
4.4% Strongly Disagree
6. Reimbursement for some methods of pain management is an issue to consider before implementing pain management practice guidelines (PMPG). 28.5% 37.2% 11.7% 13.1% 9.5% Strongly Mostly Unsure Mostly Strongly Agree Agree Disagree Disagree 7. The implementation of PMPG would improve market value to third-party payers because more efficient and appropriate patterns of care would produce the highest quality outcome. 36.5% 40.9% 19.0% 2.2% 1.5% Strongly Mostly Unsure Mostly Strongly Agree Agree Disagree Disagree 8. The implementation of PMPG would lead to more timely and effective treatment by education of patients about the need to communicate unrelieved pain. 49.6% 40.9% 8.8% 0.7% 0.0% Strongly Mostly Unsure Mostly Strongly Agree Agree Disagree Disagree 9. The risk of medical liability may actually be reduced for those practitioners who follow PMPGs. 23.4% 43.8% 28.5% 3.6% Strongly Mostly Unsure Mostly Agree Agree Disagree
0.7% Strongly Disagree
10. The implementation of PMPG would contribute to fewer postoperative complications and, in some cases, shorter stays after surgical procedures. 32.8% 46.7% 18.2% 1.5% 0.7% Strongly Mostly Unsure Mostly Strongly Agree Agree Disagree Disagree 11. The implementation of PMPG would reduce the incidence and severity of patients’ postoperative or posttraumatic pain. 55.5% 38.0% 5.1% 0.7% 0.7% Strongly Mostly Unsure Mostly Strongly Agree Agree Disagree Disagree 12. The implementation of PMPG would enhance patient comfort and satisfaction. 67.7% 27.9% 3.7% Strongly Mostly Unsure Agree Agree
0.7% Mostly Disagree
0.0% Strongly Disagree
CEO ⫽ chief executive officer.
J. Clin. Anesth., vol. 13, June 2001
271
Original Contributions
Table 3. Results of the Clinical Director Survey on Pain Management Practices (n ⫽ 135) 1. Do you have a comprehensive acute postoperative pain-management program at your hospital? 6.5% 53.2% 40.3% No Response Yes No (skip to question 3) 2. Did you use a nationally published clinical practice guideline to establish your acute postoperative pain management program? 0.0% 50% 50% No Response Yes No (n ⫽ 74 responding ‘yes’ to #1) 3. For what proportion of patients do you use a collaborative, interdisciplinary approach to acute postoperative pain control (i.e., an approach that includes all members of the health care team?) 6.5% 10.8% 30.9% 44.6% 5.0% No response All Patients Most Patients Some Patients No Patients 4. For what proportion of patients is an individualized proactive pain control plan developed preoperatively by patients and practitioners? 5.0% 8.6% 30.2% 51.1% 5.0% No response All Patients Most Patients Some Patients No Patients 5. For what proportion of patients is a complete pain history, including previous pain experiences and drug/nondrug therapies, obtained preoperatively? 5.0% 11.5% 20.9% 50.4% 12.2% No response All Patients Most Patients Some Patients No Patients 6. With what proportion of patients do you discuss patient and family members’ expectations, preferences, and beliefs about pain assessment and management? 4.3% 5.8% 25.9% 59.7% 4.3% No response All Patients Most Patients Some Patients No Patients 7. With what proportion of patients do you use a formal postoperative pain assessment tool such as a Visual Analog Scale, Descriptive scale, etc.? 5.0% 25.9% 22.3% 28.1% 18.7% No response All Patients Most Patients Some Patients No Patients 8. Are the following modalities used at your hospital for acute postoperative pain management? Yes Pharmacologic Interventions: No Response Opioids via oral 8.6% 90.6% intramuscular/subcutaneous 5.8% 94.2% intravenous 5.0% 94.2% patient-controlled analgesia (intravenous) 5.8% 84.9% epidural or Intrathecal 0.8% 79.1% Local Anesthestics epidural or Intrathecal 7.9% 82.0% peripheral nerve block 6.5% 82.7% Nonsteroidal Anti-Inflammatory Drugs (NSAIDs) oral 5.0% 93.5% parenteral 14.4% 76.3% Non-Pharmacologic Interventions: relaxation Techniques education/Instruction transcutaneous Electrical Nerve Stimulation (TENS)
Factor Analysis of the CEOs’ Responses One hundred twenty-seven hospitals responded to both the CEO survey and the clinical director survey, and were included in the factor analysis and logistic regression. Factor analysis of responses to the CEO survey revealed four underlying factors measured by the 12 survey questions of CEOs’ perceptions of issues related to guideline implementation (Table 4). For purposes of discussion, these four hypothetical factors were labeled as institutional resources, economic/legal incentives, patient out272
J. Clin. Anesth., vol. 13, June 2001
8.6% 7.2% 10.1%
33.1% 64.0% 46.8%
No 0.7% 0.0% 0.7% 9.4% 10.1% 10.1% 10.8% 1.4% 9.4% 58.3% 28.8% 43.2%
comes, and reimbursement concerns. The matrix shows how each survey question is correlated with each factor. Questions 1, 2, 4, and 5 load highly and significantly on the first common factor that reflects availability of specialty staff and equipment. Questions 7 to 9 load highly on the common factor that refers to the potential of using practice guidelines to achieve economic and legal advantages. Questions 10 to 12 load significantly on the third factor that addresses the benefits of guideline adoption to improve patient outcomes, including reduced complica-
Acute pain management guidelines: Jiang et al.
Table 4. Factor Analysisa of Hospital CEOs’ Perceptions of Issues Influencing the Adoption of Pain Management Practice Guidelines (n ⫽ 127 hospitals) FACTORS 1 Institutional Resources
2 Economic/Legal Incentives
3 Patient Outcomes
4 Reimbursement Concerns
0.7979 0.6654 0.6867 0.6441
⫺0.0235 0.2748 ⫺0.1789 0.1430
⫺0.0782 ⫺0.1189 0.1777 0.0589
⫺0.0183 0.0015 0.0118 ⫺0.0556
0.0188 0.0463 0.0577
0.7757 0.5874 0.7889
0.1956 0.4153 0.1564
0.0673 0.1663 ⫺0.1134
#10 Complications #11 Severity of pain #12 Patient satisfaction
0.0889 ⫺0.0158 0.0071
0.4903 0.2708 0.1629
0.5513 0.8679 0.8635
0.1726 ⫺0.0290 ⫺0.1135
#3 Use of epidural catheters #6 Reimbursement for some pain management methods
0.4853 ⫺0.2602
⫺0.1235 0.1522
0.0723 ⫺0.0924
0.6106 ⴚ0.8193
Questions on CEO Surveys #1 #2 #4 #5
Training of hospital staff Equipment Staff for non-drug treatment Formal QA program
#7 Market value to payers #8 Timely, effective treatment #9 Medical liability
a The factor matrix shows the correlations between the original variables and the derived factors. These four factors explain 63.4% of the total variance in the [chief executive officer] (CEO) survey.
tions and pain acuity and increased patient satisfaction. Questions 3 and 6 are related to the fourth factor that concerns reimbursement for some pain management methods as recommended by guidelines. It is noted that Question 10 loads significantly on both Factors 2 and 3, suggesting that reduced postoperative complications not only reflect improved patient outcomes but also have significant legal implications. Question 3 loads significantly on both Factors 1 and 4, suggesting that if reimbursement is a concern for some pain management regimens, expertise may not be available at hospitals to deliver these pain treatments. In summary, the four extracted factors account for 63.4% of the total variance in the CEO survey, which is considered to be a satisfactory solution.
Prediction of Guideline Implementation Results of the logistic regression show how a hospital CEO’s perception of factors affecting guideline implementation was related to the likelihood of the hospital’s adoption of a pain management guideline (Table 5). After controlling for hospital characteristics (i.e., bed size, teaching status, urban/rural location, and sponsorship), three factors were found to be significantly associated with guideline implementation: 1) availability of institutional resources; 2) belief in the favorable impact of guideline usage on patient outcomes; and 3) belief in achieving economic and legal benefits through guideline adoption. Both availability of institutional resources and the belief in improved patient outcomes were strong predictors of guideline implementation. Reimbursement concerns, on the other hand, were not significantly related to the likelihood of adopting a pain management guideline. Logistic regression analysis was used further to identify those factors, as perceived by hospital CEOs, that signifi-
cantly predict the likelihood of complying with key elements of the AHRQ/AHCPR guideline. The results are summarized in Table 6. Availability of institutional reTable 5. Relationship Between the Likelihood of Adopting a Pain Management Guideline and the Hospital CEO’s Perception of Issues Related to Guideline Implementationa 95% Confidence Odds Ratio Interval p-valueb
Independent Variables Hospital CEOs’ perception on issues related to guideline implementationc Institutional resources Economic/legal incentives Patient outcomes Reimbursement concerns Control variables:d Bed size Teaching Urban Voluntary
2.433 1.668 2.526 1.160
(1.417,4.179) (1.025,2.714) (1.323,4.824) (0.731,1.838)
0.0013 0.0394 0.0050 0.5287
1.002 0.551 1.076 1.151
(1.000,1.004) (0.161,1.882) (0.251,4.615) (0.265,5.008)
0.1053 0.3416 0.9215 0.8509
a
Logistic regression model was estimated with the dependent variable defined as 1 if the hospital reported using a nationally published pain management practice guideline and 0 if not. b
p-value: those underscored are significant at the 0.05 level. Hospital CEOs’ perception on issues related to guideline implementation was measured by four composite variables (called factors) that were derived from the 12 questions on the CEO survey (see Table 4). c
d
Teaching status was measured as 1 for teaching hospitals and 0 for nonteaching hospitals. Urban was measured as 1 for urban location and 0 for the rural. Over 95% of the community hospitals in New York State are nonprofit. Sponsorship was measured as 1 for voluntary and 0 for public. J. Clin. Anesth., vol. 13, June 2001
273
Original Contributions
Table 6. Relationship Between the Likelihood of Complying with Key Elements of the AHRQ/AHCPR Acute Pain Management Guideline and the Hospital CEO’s Perception of Issues on Guideline Implementationa Key Elements of the AHRQ/AHCPR Guideline Reported Being Used for Most to All Surgical Inpatients
Hospital CEO’s Perceptions (Independent Factors)
Interdisciplinary Team
Institutional resources Economic/legal incentives Patient outcomes Reimbursement concerns
2.795* (1.662, 1.185 (0.788, 1.266 (0.811, 0.893 (0.592,
4.701) 1.782) 1.976) 1.345)
Pain Control Plan Developed Preoperatively
Documentation of Pain History
Discussion with Patients and Families
Use of Formal Pain Assessment Tool
2.685* (1.640, 4.396) 0.957 (0.642, 1.425) 1.413 (0.920, 2.170) 1.602‡ (1.044, 2.457)
2.449* (1.490,4.022) 0.949 (0.631,1.429) 1.075 (0.710,1.627) 1.502 (0.965,2.337)
2.065† (1.275, 3.345) 1.036 (0.690, 1.556) 1.387 (0.889, 2.163) 1.556‡ (1.005, 2.410)
2.375* (1.500,3.760) 1.232 (0.829,1.833) 1.176 (0.779,1.775) 1.263 (0.843,1.894)
a
Logistic regression model was estimated for each key element, with the dependent variable defined as 1 if the hospital reported implementing this key element for most to all surgical inpatients, and 0 if not. Odds ratio and 95% confidence intervals (CIs)are presented. Each model has been controlled for hospital characteristics including bed size, teaching status, urban/rural location, and sponsorship. AHRQ/AHCPR ⫽ Agency for Healthcare Research and Quality/Agency for Healthcare Policy and Research, CEO⫽chief executive officer. *p ⬍ 0.001. p ⬍ 0.01. p ⬍ 0.05.
† ‡
sources was found to be a significant predictor for compliance with five key elements of the AHRQ/AHCPR guideline: use of an interdisciplinary team, development of a pain control plan preoperatively, documentation of the patient’s pain history, discussion with patients and families about pain management, and use of a formal pain assessment tool. The results also indicate that when reimbursement was considered an issue, a hospital was more likely to have an individualized pain control plan developed preoperatively and include patients and families in the discussion. Given the importance of resource availability to guideline implementation, additional analysis was conducted to examine the association of each resource aspect with use of guidelines and compliance of five key elements of the AHRQ/AHCPR guideline (Table 7). First, hospitals were stratified into two groups based on their CEOs’ responses to each of five questions concerning resource availability; that is, whether the hospital CEO mostly/strongly agreed or mostly/strongly disagreed with each resource statement. Then a comparison was made in the percentage of hospitals reporting using a guideline between the two groups of hospitals (i.e., “agree” vs. “disagree” on each resource aspect). Clinical directors reported using a pain management guideline significantly more often in hospitals where staff are trained and knowledgeable in pain management, particularly nondrug intervention. Similar findings were shown for compliance with key elements of the AHRQ/ AHCPR guideline. In particular, existence of a formal QA program to monitor pain management was significantly associated with use of a pain management guideline and compliance with several key elements of the AHRQ/ AHCPR guideline. Availability of enough specialists for epidural catheters was significantly associated with hospitals complying with two key elements: development of a 274
J. Clin. Anesth., vol. 13, June 2001
pain control plan preoperatively and inclusion of the patients and families in the formulation of the plan.
Discussion Based on the survey results, a majority of hospital CEOs agreed that their clinical staff were generally trained and knowledgeable in pain management, and that they had the equipment necessary to deliver the recommended interventions. Most of the surveyed CEOs also believed that implementing a pain management practice guideline was beneficial in terms of lowered cost, improved quality of care, and reduced risk of medical liability. Nevertheless, only 27% of the surveyed hospitals reported using a nationally published clinical practice guideline. Less than 50% of the respondents said that their current pain management programs contained those key elements as recommended by the AHRQ/ AHCPR guideline such as use of an interdisciplinary approach and an individualized proactive pain control plan. There also appeared to be a lack of staff trained in delivering nondrug interventions and an under-utilization of nonpharmacological therapies to control pain. Further analysis of the survey results using factor analysis and logistic regression suggested that three major factors were significantly associated with the implementation of pain management practice guidelines. The first factor was availability of institutional resources, particularly the required personnel for pain management and the existence of a formal QA program or other institutional program to monitor pain management. These findings confirm some of the concerns raised by Carr and colleagues regarding local resource shortfalls for implementing pain management practice guidelines.22 The second impetus for hospitals to adopt pain management guidelines was to improve patient outcomes including reduced
Acute pain management guidelines: Jiang et al.
Table 7. Association of Guideline Usage with Hospital CEOs’ Perception of Resource Availability
Resource Availability (based on the CEOs’ response to each of the following statements) Hospital staff is trained and knowledgeable in pain management –Agree –Disagree Hospital has the equipment to deliver various pain treatment –Agree –Disagree Hospital has enough specialists with expertise to do epidural catheters –Agree –Disagree Hospital has enough trained staff to deliver non-drug interventions –Agree –Disagree Hospital has a formal QA program to monitor pain management –Agree –Disagree
Key Elements of the AHRQ/AHCPR Guideline Reported Being Used for Most to All Surgical Inpatients (% who reported implementing for most to all inpatient surgical patients)
Use of a Pain Management Guideline (% Interdisciplinary reporting “Yes”) Team
Developing Pain Control Plan
Documentation of Pain History
Discussion With Patients/ Families
Use of Pain Assessment Tool
33.3% 11.8% (p ⫽ 0.016)
52.1% 17.1% (p ⫽ 0.000)
45.8% 28.6% (p ⫽ 0.076)
40.6% 14.3% (p ⫽ 0.005)
40.6% 14.3% (p ⫽ 0.005)
55.2% 34.3% (p ⫽ 0.034)
29.8% 7.7% (p ⫽ 0.091)
44.9% 23.1% (p ⫽ 0.131)
43.2% 23.1% (p ⫽ 0.161)
33.1% 38.5% (p ⫽ 0.695)
33.1% 38.5% (p ⫽ 0.695)
53.4% 15.4% (p ⫽ 0.009)
29.1% 17.6% (p ⫽ 0.326)
45.1% 27.8% (p ⫽ 0.167)
45.1% 16.7% (p ⫽ 0.023)
35.4% 22.2% (p ⫽ 0.272)
37.2% 11.1% (p ⫽ 0.030)
52.2% 33.3% (p ⫽ 0.137)
41.5% 20.9% (p ⫽ 0.015)
55.8% 36.4% (p ⫽ 0.035)
60.5% 31.8% (p ⫽ 0.002)
60.5% 20.5% (p ⫽ 0.000)
55.8% 22.7% (p ⫽ 0.000)
55.8% 46.6% (p ⫽ 0.321)
33.7% 14.6% (p ⫽ 0.024)
47.2% 33.3% (p ⫽ 0.135)
48.3% 26.2% (p ⫽ 0.016)
40.4% 19.0% (p ⫽ 0.015)
38.2% 23.8% (p ⫽ 0.104)
58.4% 31.0% (p ⫽ 0.003)
p-value underlined denotes that it is significant at the 0.05 level. CEO ⫽ chief executive officer, AHRQ/AHCPR ⫽ Agency for Healthcare Research and Quality/Agency for Healthcare Policy and Research, QA ⫽ quality assurance.
pain severity and enhanced patient satisfaction. The third driving force for guideline implementation related to economic and legal incentives such as enhanced market value to payers and reduced medical liability. The potential of reducing medical liability through guideline implementation derives from the premise that practice guidelines are built on evidence-based medicine and, therefore, can justify or assist physicians in their clinical decision making.26 Apparently, the first major factor determining the implementation of pain management practice guidelines was a hospital’s internal resources, while the second and third factors reflected individual drive toward clinical excellence and expectations from a variety of external audiences, including payers, consumers, and regulatory entities. Although 65% of the responding CEOs considered reimbursement for some pain management methods an issue for implementing pain management practice guidelines, this factor was not found to be significantly associated with the adoption of guidelines. One possible explanation is that when considered together with all other
factors in the logistic regression analysis, the influence of this variable became relatively small. Concerns for reimbursement of certain recommended interventions did not seem to impede the use of pain management practice guidelines. Nevertheless, the results of this study also indicate that when reimbursement was considered an issue, hospitals were less likely to have expertise available to deliver some specialized pain management regimens. Facilitating consistent payment criteria is therefore important to support appropriate use of all available pain treatment methods. Several limitations of this study need to be addressed. First, our survey sample includes only hospitals located in New York State, which may limit the generalizability of our findings to hospitals in other states. Conversely, New York hospitals represent a wide variety of providers, including major teaching hospitals located in large metropolitan areas such as New York City and small community hospitals serving rural or mountainous communities in the upstate area. This situation allowed us to observe certain J. Clin. Anesth., vol. 13, June 2001
275
Original Contributions
variations across different hospitals. Second, questions on the CEO survey that related to availability of institutional resources were only proxy measures based on the CEO’s perception. Perhaps it would have been more accurate to measure the size and composition of the pain management staff, the availability of equipment, and other organizational attributes. Third, the CEO survey may not have included all factors affecting the implementation of pain management guidelines. These factors could include the existence of regulatory obstacles that may inhibit appropriate pain management, and the influence of accreditation bodies such as the JCAHO. In conclusion, the findings of this study contribute to the understanding of factors influencing the implementation of pain management practice guidelines in hospital settings. The external environment can generate a variety of incentives for using pain management practice guidelines, but internal resource shortfalls may significantly inhibit implementation. In light of the fact that pain management has recently been incorporated in the JCAHO accreditation process, this information is both timely and important for both clinicians and health care executives. Many hospitals may consider implementing pain management practice guidelines as part of their efforts to comply with the JCAHO standards. To ensure the successful implementation of such guidelines, it is important for hospitals to carefully assess their internal resources, including the staff’s knowledge and skills in pain management, and to establish an institutional approach that results in a measurable improvement in outcomes. Implementing practice guidelines can contribute to reduced pain severity and enhanced patient satisfaction. Strategies are still needed for developing other effective mechanisms to support and encourage desired practice. Ultimately, real improvements in diagnosis and treatment of pain will require a coordinated effort of public policy, providers, patients, and payers.
Acknowledgments The authors wish to acknowledge Onita Munshi, M.S., MBA, a former researcher of HANYS, for her indispensable assistance with survey construction and administration, data collection, and analysis at the early stage of this project.
References 1. Pain terms: a list with definitions and notes on usage. Recommened by the IASP. Pain 1979;6:249. 2. Abbott FV, Gray-Donald K, Sewitch MJ, Johnston CC, Edgar L, Jeans ME: The prevalence of pain in hospitalized patients and resolution over six months. Pain 1992;50:15–28. 3. Donovan M, Dillon P, McGuire L: Incidence and characteristics of pain in a sample of medical-surgical inpatients. Pain 1987;30: 69 –78. 4. Lynch EP, Lazor MA, Gellis JE, Orav J, Goldman L, Marcantonio ER: Patient experience of pain after elective noncardiac surgery. Anesth Analg 1997;85:117–123. 5. Oden R: Acute postoperative pain: incidence, severity, and the etiology of inadequate treatment. Anesth Clin N Am 1989;7:1–15. 6. Agency for Healthcare Research and Quality, formerly Agency for Health Care Policy and Research: Acute pain management: operative or
276
J. Clin. Anesth., vol. 13, June 2001
7.
8.
9.
10.
11. 12. 13.
14. 15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
medical procedures and trauma. AHRQ Publication No. 92–0032. Rockville, MD: US Department of Health and Human Services, 1992. Am Pain Society: Am Pain Society quality assurance standards for relief of acute pain and cancer pain. In: Bond MR, Charlton JE, Woolf CJ (eds): Proceedings of the VIth World Congress on Pain. New York: Elsevier Science Publishers, 1991. Practice guidelines for acute pain management in the perioperative setting. A report by the Am Society of Anesthesiologist task force on pain management, acute pain section. Anesthesiology 1995;82:1071– 81. Bach DM: Implementation of the Agency for Health Care Policy and Research postoperative pain management guideline. Nurs Clin North Am 1995;30:515–27. Bookbinder M, Coyle N, Kiss M, et al: Implementing national standards for cancer pain management: program model and evaluation. J Pain Symptom Manage 1996;12:334 –347. Campese C: Development and implementation of a pain management program. AORN J 1996;64:937– 40. Knowles R: Standardization of pain management in the postanesthesia care unit. J Perianesths Nurs 1996;11:390 – 8. Davis DA, Taylor-Vaisey A: Translating guidelines into practice. A systematic review of theoretic concepts, practical experience and research evidence in the adoption of clinical practice guidelines. CMAJ 1997;157:408 –16. Rogers EM: Lessons for guidelines from the diffusion of innovations. Jt Comm J Qual Improv 1995;21:324 – 8. James PA, Cowan TM, Graham RP, Majeroni BA: Family physicians’ attitudes about and use of clinical practice guidelines. J Fam Pract 1997;45:341–7. Lawler FH, Viviani N: Patients and physician perspectives regarding treatment of diabetes: Compliance with practice guidelines. J Fam Pract 1997;44:369 –73. Weingarten S, Stone E, Hayward R, et al: The adoption of preventive care practice guidelines by primary care physicians: do actions match intentions? Gen Intern Med 1995;10:138 – 44. Reis SE, Holubkov R, Zell KA, Edmundowicz D, Shapiro AH, Feldman AM: Unstable angina: specialty-related disparities in implementation of practice guidelines. Clin Cardiol 1998;21:207–10. Ely JW, Goerdt CJ, Bergus GR, West CP, Dawson JD, Doebbeling BN: The effect of physician characteristics on compliance with adult preventive care guidelines. Fam Med 1998;30:34 –9. Goodpastor WA, Montoya ID: Motivating physician behavior change: Social influence versus financial contingencies. Int J Health Care Qual Assur 1996;9:4 –9. Petrazzuoli M, Rothe MJ, Grin-Jorgensen C, Ramsey WH, GrantKels JM: Monitoring patients taking methotrexate for hepatotoxicity. Does the standard of care match published guidelines? J Am Acad Dermatol 1994;31:969 –77. Carr D, Jacox A, Chapman C, et al.: Acute pain management: guideline technical report (No.1). Rockville, MD: Agency for Health Care Policy and Research, Public Health Service, U.S. Department of Health and Human Services; 1995. Hair JF Jr., Anderson RE, Tatham RL, Black WC: Multivariate Data Analysis with Readings. New York, NY: Macmillan Publishing Company, 1992. Ware JE Jr, Kosinski M, Gandek B, et al: The factor structure of the SF-36 Health Survey in 10 countries: results from the IQOLA Project. International Quality of Life Assessment. J Clin Epidemiol 1998;51:1159 – 65. Zaslavsky AM, Beaulieu ND, Landon BE, Cleary PD: Dimensions of consumer-assessed quality of Medicare managed-care health plans. Med Care 2000;38:162–74. Merritt TA, Palmer D, Bergman DA, Shiono PH: Clinical practice guidelines in pediatric and newborn medicine: implications for their use in practice. Pediatrics 1997;99:100 –14.
*Senior Researcher, Healthcare Association of New York State, Albany, NY. Currently Social Scientist, Agency for Healthcare Research and Quality, Rockville, MD. †Associate Professor of Clinical Anesthesiology, Montefiore Medical Center, Bronx, NY ‡Vice President, Division of Quality & Research Initiatives, Healthcare Association of New York State, Albany, NY §Vice President, Medical Affairs, Ellis Hospital, Schenectady, NY 储Chairman of Anesthesiology, Long Island Jewish Medical Center, New Hyde Park, NY Address correspondence and reprint requests to Dr. Lagasse at the Department of Anesthesiology, Montefiore Medical Center – Weiler Division, 1825 Eastchester Rd., Bronx, NY 10461. E-mail: [email protected] This study was part of a larger project that was funded by the New York State Department of Health, Albany, NY, as recommended by the Task Force on Clinical Guidelines and Technology Assessment (Grant #TF050). Received for publication October 4, 2000; revised manuscript accepted for publication March 12, 2001.
Study Objective: To identify factors that may influence the implementation of acute pain management guidelines in hospital settings. Design: Two questionnaire surveys. Setting: Healthcare Association of New York State, Albany, NY. Measurement: The surveys were administered to 220 hospitals in New York State regarding their acute pain management practices and resources available. One survey was addressed to each hospital’s chief executive officer (CEO) and the second survey was addressed to the clinical director of the Department of Anesthesiology or Acute Pain Service. The barriers and incentives to guideline implementation identified by CEOs were analyzed using factor analysis. Logistic regression was employed to determine predictors of guideline implementation by linking the CEOs’ survey data with the clinical directors’ report of guideline usage. Main Results: According to clinical directors, only 27% of the responding hospitals were using a published pain management practice guideline. Factors predictive of guideline implementation include resource availability and belief in the benefits of using guidelines to improve quality of care or to achieve economic/legal advantages. Guideline implementation, however, does not necessarily include applying all key elements recommended by the federal Agency for Healthcare Research and Quality (formerly Agency for Health Care Policy and Research) guideline. For example, a collaborative, interdisciplinary approach to pain control was used in only 42% of the hospitals, and underutilization of nonpharmacologic therapies to control pain was widespread. Resource availability, particularly staff with expertise in pain management and existence of a formal quality assurance program to monitor pain management, was significantly predictive of compliance with key guideline elements. Conclusions: Resource availability significantly influences the implementation of pain management practice guidelines in hospital settings. Implementation is often incomplete because various factors affect the feasibility of individual guideline elements and may explain the varying results that guidelines have had on clinical practices. © 2001 by Elsevier Science Inc. Keywords: Acute pain management, clinical practice guidelines, quality improvement.
Journal of Clinical Anesthesia 13:268 –276, 2001 © 2001 Elsevier Science Inc. All rights reserved. 655 Avenue of the Americas, New York, NY 10010
0952-8180/01/$–see front matter PII S0952-8180(01)00268-9
Acute pain management guidelines: Jiang et al.
Introduction Pain is described as “an unpleasant sensory and emotional experience associated with actual or potential tissue damage.”1 Inadequate treatment of postoperative and other acute pain is well documented in the medical literature.2–5 In response, several national entities, such as the federal Agency for Healthcare Research and Quality (AHRQ; formerly Agency for Health Care Policy and Research, AHCPR), the American Pain Society (APS), and the American Society of Anesthesiologists (ASA), have developed pain management practice guidelines to facilitate better assessment and treatment of pain.6 – 8 Although implementation of these practice guidelines was found to have a favorable impact on improving patient outcomes and satisfaction,9 –12 the guidelines have not been adopted extensively at the local level. The importance of addressing pain management for improving quality of care is well reflected by the current initiative of the Joint Commission on Accreditation of Healthcare Organizations (JCAHO), which will use guideline-based measures of pain management in its survey process in the year 2001. This fact presents strong incentives for hospitals and other health care organizations to adopt pain management practice guidelines. Inevitably, there will be increasing interest in what influences the implementation process and how to ensure the successful adoption of guidelines. A vast amount of research has been conducted to examine factors that affect the implementation of clinical practice guidelines.13 Variables that have been identified as important to guideline implementation include: 1) attributes of guidelines14; 2) physician characteristics, including attitudes about guidelines15–17; specialty training18; and gender19; 3) practice setting15,20; 4) patient factors21; and 5) incentives related to financial, legal, and regulatory issues. Nevertheless, to date, few empirical studies have addressed barriers to the adoption of pain management practice guidelines. Most guidelines give little attention to the feasibility of implementing the recommendations across a wide variety of health care settings. In reviewing the AHRQ/AHCPR acute pain management guideline, Carr and colleagues suggested that the following local resource shortfalls might limit implementation22: 1) lack of staff trained and knowledgeable in pain management, 2) lack of specially trained nurses or psychologists to deliver nondrug interventions, 3) lack of formal quality assurance or other institutional programs to monitor pain management, 4) absence of equipment necessary to deliver some of the interventions, 5) lack of specialists with the requisite expertise to insert and monitor epidural catheters, and 6) lack of resources to provide the education to clinicians to implement the guideline. Prior studies on clinical practice guidelines tend to focus on physicians’ perspectives on guideline implementation, while little is known about health executives’ perspectives on barriers to guideline adoption. Good understanding and support from hospital administration is imperative to the successful implementation of acute pain management practice guidelines. This study seeks to
facilitate understanding of select issues related to implementation of a clinical practice guideline on acute pain management. Three questions were addressed through surveys of hospital chief executive officers (CEOs) and clinical directors. First, what are the hospital CEOs’ perceptions of factors influencing pain management guideline implementation? Second, how do current pain management practices, as reported by the clinical directors, compare to what is recommended by the guideline? Third, do hospital CEOs’ perceptions of issues related to guideline implementation predict the adoption of guideline elements as reported by the clinical directors?
Materials and Methods To identify factors that may influence the decision by a hospital to implement acute pain management practice guidelines, two sets of surveys were developed, pilot tested, and mailed to 220 hospitals in New York State. One survey was addressed to each hospital’s CEO and the second survey was addressed to the clinical director of the Departments of Anesthesiology or Acute Pain Service. The CEOs were asked the extent to which they agree or disagree with statements about the accessibility of staff trained and knowledgeable in all modalities of pain management, availability of equipment necessary to deliver the recommended interventions, reimbursement issues, and possible benefits to the hospital of implementing a pain management guideline. The clinical directors were asked about the type of acute postoperative pain management program being instituted, the use of a nationally published clinical practice guideline, and current pain management practices. Questions on current pain management practices assessed compliance with elements of the AHRQ/ AHCPR acute pain management practice guideline. Responses on both surveys were graded on a five-point Likert scale ranging from “strongly agree” to “strongly disagree.” Responses to the CEO survey were summarized using factor analysis to identify underlying dimensions (called “factors”) that were measured by the 12 survey questions. Factor analysis is a multivariate statistical technique that analyzes the interrelationship among a large number of variables and summarizes these variables in terms of their common underlying factors.23 It allows one to generate a more parsimonious description of the survey results. Factor analysis has been widely used in health services research for validating survey instruments. For example, Ware and colleagues at the New England Medical Center used factor analysis to validate the eight physical and mental health components measured by the SF-36 Health Survey.24 Zaslavsky and colleagues at Harvard Medical School used factor analysis to summarize the 39-item Consumer Assessment of Health Plans Survey (CAHPS) into four subjects, including interactions with clinical staff, customer service, access to care, and advice on healthpromoting activities.25 The second purpose of performing factor analysis in this study was to reduce the original 12 survey questions to a smaller number of factors that subsequently would be used in the logistic regression J. Clin. Anesth., vol. 13, June 2001
269
Original Contributions
Table 1. Descriptive Statistics for Hospital Characteristics
Hospital Characteristics Bed size ⬍ 250 beds 250–499 beds 500 ⫹ beds Teaching status Major teaching Minor teaching Nonteaching Location Urban Rural Sponsorship Voluntary Public Proprietary
Clinical CEO Director Survey Survey Respondents Respondents All New York (nⴝ137) (nⴝ135) State Hospitals
49.6% 32.8% 17.5%
49.2% 34.1% 16.7%
55.8% 30.4% 13.8%
27.0% 24.8% 48.2%
25.8% 25.0% 49.2%
23.1% 26.9% 50.0%
83.9% 16.1%
83.3% 16.7%
84.2% 15.8%
90.5% 9.5% –
90.2% 9.8% –
85.8% 10.8% 3.5%
CEO ⫽ chief executive officer.
analysis. Each factor represents a composite of all the variables related to that particular factor. Logistic regression analysis was conducted to validate the significance of hospital CEOs’ perceptions of factors related to guideline implementation by using their responses to predict the usage of pain management guidelines as reported by the clinical directors. The outcome variable was set equal to 1 if the clinical director reported using a guideline, and equal to 0 otherwise. The model is: log[P/(1 ⫺ P)] ⫽ 0 ⫹ 1X ⫹ 2Y, where P is the probability of implementing a guideline, and X is a vector of hypothetical factors derived from the factor analysis of the CEO survey, while Y is a vector of hospital characteristics including bed size, teaching status, urban/rural location, and sponsorship. The results of the logistic regression were used to identify factors significantly related to the implementation of a pain management practice guideline. Similarly, logistic regression analysis was used to further identify those factors, as perceived by hospital CEOs, that significantly predict the likelihood of complying with key elements of the AHRQ/AHCPR guideline.
Results The results are presented in three subsections. The first subsection describes the results of the CEO survey and the clinical director survey. The second subsection reports how the results of the 12 questions on the CEO survey were summarized into four areas of perceived barriers/ incentives to guideline implementation: a) availability of institutional resources, b) economic/legal incentives for guideline implementation, c) potential impact on patient outcomes, and d) concerns on reimbursement. The last subsection describes how a hospital CEO’s perception 270
J. Clin. Anesth., vol. 13, June 2001
of issues on guideline implementation was related to the actual adoption of pain management practice guidelines in the hospital and compliance of key guideline elements.
The CEO and the Clinical Director Surveys The surveys were administered to 220 hospitals in New York State in 1996 and 1998 as part of a three-year study funded by the New York State Department of Health to evaluate the impact of pain management practice guideline implementation on clinical practices and patient outcomes. One hundred thirty-seven (62.3% response rate) of the CEOs and 135 (61.4%) of the clinical directors responded to the first round of surveys in 1996 while only 90 hospitals responded to the second round of surveys in 1998. No significant difference was found between the results of the second round of surveys and those of the initial surveys. Therefore, we chose to present the results of the first round of surveys which had a larger hospital sample size. Hospital characteristics are summarized in Table 1. There was no statistically significant difference in any of the characteristics between the survey sample and the New York State hospital population, suggesting that the survey sample well represented New York hospitals. The results of the CEO survey are presented in Table 2. One hundred three (74%) of the CEOs responding agreed that their staff were trained and knowledgeable in pain management, but only 47 (34%) indicated that they had enough specially trained staff to deliver nondrug interventions. One hundred twenty-three (90%) CEOs also agreed that they had the necessary equipment. A majority of the responding hospitals (69%) had formal quality assurance (QA) programs to monitor pain management. The CEOs agreed that implementation would improve market value to payers (77%), risk of medical liability may be reduced for practitioners following the guideline (67%), and that patients would have fewer complications, and in some cases, shorter hospital stays (80%). Eighty-nine (65%) CEOs, however, agreed that reimbursement was an issue to consider before implementing pain management practice guidelines. Of the clinical directors responding to the survey (Table 3), 76 (56%) stated that they had a comprehensive acute pain management program, but only 37 (27%) used a published clinical practice guideline to establish their program. A collaborative, interdisciplinary approach to pain control was used for all or most inpatients in 42% of the hospitals responding. Individualized proactive pain control plans were developed preoperatively in all or most inpatients in 39% of the hospitals. Close to 50% of the hospitals reported using a formal postoperative pain assessment tool. Although 89% to 100% of the hospitals used pharmacologic interventions such as opioids, local anesthetics, and nonsteroidal antiinflammatory drugs (NSAIDs), the percentage of hospitals offering nonpharmacologic interventions such as patient education (69%), transcutaneous electrical nerve stimulation (TENS) (52%), and relaxation therapy (36%), was considerably lower.
Acute pain management guidelines: Jiang et al.
Table 2. Results of the Hospital CEO Survey (n ⫽ 137) 1. Our hospital staff (including physicians, nurses, psychologists, and others) is generally trained and knowledgeable in pain management. 18.4% 55.1% 5.9% 19.9% 0.7% Strongly Mostly Unsure Mostly Strongly Agree Agree Disagree Disagree 2. Our hospital has the equipment necessary to deliver some of the interventions for pain relief (including various drugs and pumps for patient-controlled analgesia.) 59.1% 31.4% 0.7% 7.3% 1.5% Strongly Mostly Unsure Mostly Strongly Agree Agree Disagree Disagree 3. Our hospital has enough specialists with the requisite expertise to insert and monitor epidural catheters. 57.4% 28.7% 1.5% 9.6% Strongly Mostly Unsure Mostly Agree Agree Disagree
2.9% Strongly Disagree
4. Our hospital has enough specially trained nurses and psychologists to deliver nondrug interventions. 8.1% 25.7% 30.1% 26.5% Strongly Mostly Unsure Mostly Agree Agree Disagree
9.6% Strongly Disagree
5. Our hospital has a formal quality assurance or other institutional program to monitor pain management. 29.2% 39.4% 10.9% 16.1% Strongly Mostly Unsure Mostly Agree Agree Disagree
4.4% Strongly Disagree
6. Reimbursement for some methods of pain management is an issue to consider before implementing pain management practice guidelines (PMPG). 28.5% 37.2% 11.7% 13.1% 9.5% Strongly Mostly Unsure Mostly Strongly Agree Agree Disagree Disagree 7. The implementation of PMPG would improve market value to third-party payers because more efficient and appropriate patterns of care would produce the highest quality outcome. 36.5% 40.9% 19.0% 2.2% 1.5% Strongly Mostly Unsure Mostly Strongly Agree Agree Disagree Disagree 8. The implementation of PMPG would lead to more timely and effective treatment by education of patients about the need to communicate unrelieved pain. 49.6% 40.9% 8.8% 0.7% 0.0% Strongly Mostly Unsure Mostly Strongly Agree Agree Disagree Disagree 9. The risk of medical liability may actually be reduced for those practitioners who follow PMPGs. 23.4% 43.8% 28.5% 3.6% Strongly Mostly Unsure Mostly Agree Agree Disagree
0.7% Strongly Disagree
10. The implementation of PMPG would contribute to fewer postoperative complications and, in some cases, shorter stays after surgical procedures. 32.8% 46.7% 18.2% 1.5% 0.7% Strongly Mostly Unsure Mostly Strongly Agree Agree Disagree Disagree 11. The implementation of PMPG would reduce the incidence and severity of patients’ postoperative or posttraumatic pain. 55.5% 38.0% 5.1% 0.7% 0.7% Strongly Mostly Unsure Mostly Strongly Agree Agree Disagree Disagree 12. The implementation of PMPG would enhance patient comfort and satisfaction. 67.7% 27.9% 3.7% Strongly Mostly Unsure Agree Agree
0.7% Mostly Disagree
0.0% Strongly Disagree
CEO ⫽ chief executive officer.
J. Clin. Anesth., vol. 13, June 2001
271
Original Contributions
Table 3. Results of the Clinical Director Survey on Pain Management Practices (n ⫽ 135) 1. Do you have a comprehensive acute postoperative pain-management program at your hospital? 6.5% 53.2% 40.3% No Response Yes No (skip to question 3) 2. Did you use a nationally published clinical practice guideline to establish your acute postoperative pain management program? 0.0% 50% 50% No Response Yes No (n ⫽ 74 responding ‘yes’ to #1) 3. For what proportion of patients do you use a collaborative, interdisciplinary approach to acute postoperative pain control (i.e., an approach that includes all members of the health care team?) 6.5% 10.8% 30.9% 44.6% 5.0% No response All Patients Most Patients Some Patients No Patients 4. For what proportion of patients is an individualized proactive pain control plan developed preoperatively by patients and practitioners? 5.0% 8.6% 30.2% 51.1% 5.0% No response All Patients Most Patients Some Patients No Patients 5. For what proportion of patients is a complete pain history, including previous pain experiences and drug/nondrug therapies, obtained preoperatively? 5.0% 11.5% 20.9% 50.4% 12.2% No response All Patients Most Patients Some Patients No Patients 6. With what proportion of patients do you discuss patient and family members’ expectations, preferences, and beliefs about pain assessment and management? 4.3% 5.8% 25.9% 59.7% 4.3% No response All Patients Most Patients Some Patients No Patients 7. With what proportion of patients do you use a formal postoperative pain assessment tool such as a Visual Analog Scale, Descriptive scale, etc.? 5.0% 25.9% 22.3% 28.1% 18.7% No response All Patients Most Patients Some Patients No Patients 8. Are the following modalities used at your hospital for acute postoperative pain management? Yes Pharmacologic Interventions: No Response Opioids via oral 8.6% 90.6% intramuscular/subcutaneous 5.8% 94.2% intravenous 5.0% 94.2% patient-controlled analgesia (intravenous) 5.8% 84.9% epidural or Intrathecal 0.8% 79.1% Local Anesthestics epidural or Intrathecal 7.9% 82.0% peripheral nerve block 6.5% 82.7% Nonsteroidal Anti-Inflammatory Drugs (NSAIDs) oral 5.0% 93.5% parenteral 14.4% 76.3% Non-Pharmacologic Interventions: relaxation Techniques education/Instruction transcutaneous Electrical Nerve Stimulation (TENS)
Factor Analysis of the CEOs’ Responses One hundred twenty-seven hospitals responded to both the CEO survey and the clinical director survey, and were included in the factor analysis and logistic regression. Factor analysis of responses to the CEO survey revealed four underlying factors measured by the 12 survey questions of CEOs’ perceptions of issues related to guideline implementation (Table 4). For purposes of discussion, these four hypothetical factors were labeled as institutional resources, economic/legal incentives, patient out272
J. Clin. Anesth., vol. 13, June 2001
8.6% 7.2% 10.1%
33.1% 64.0% 46.8%
No 0.7% 0.0% 0.7% 9.4% 10.1% 10.1% 10.8% 1.4% 9.4% 58.3% 28.8% 43.2%
comes, and reimbursement concerns. The matrix shows how each survey question is correlated with each factor. Questions 1, 2, 4, and 5 load highly and significantly on the first common factor that reflects availability of specialty staff and equipment. Questions 7 to 9 load highly on the common factor that refers to the potential of using practice guidelines to achieve economic and legal advantages. Questions 10 to 12 load significantly on the third factor that addresses the benefits of guideline adoption to improve patient outcomes, including reduced complica-
Acute pain management guidelines: Jiang et al.
Table 4. Factor Analysisa of Hospital CEOs’ Perceptions of Issues Influencing the Adoption of Pain Management Practice Guidelines (n ⫽ 127 hospitals) FACTORS 1 Institutional Resources
2 Economic/Legal Incentives
3 Patient Outcomes
4 Reimbursement Concerns
0.7979 0.6654 0.6867 0.6441
⫺0.0235 0.2748 ⫺0.1789 0.1430
⫺0.0782 ⫺0.1189 0.1777 0.0589
⫺0.0183 0.0015 0.0118 ⫺0.0556
0.0188 0.0463 0.0577
0.7757 0.5874 0.7889
0.1956 0.4153 0.1564
0.0673 0.1663 ⫺0.1134
#10 Complications #11 Severity of pain #12 Patient satisfaction
0.0889 ⫺0.0158 0.0071
0.4903 0.2708 0.1629
0.5513 0.8679 0.8635
0.1726 ⫺0.0290 ⫺0.1135
#3 Use of epidural catheters #6 Reimbursement for some pain management methods
0.4853 ⫺0.2602
⫺0.1235 0.1522
0.0723 ⫺0.0924
0.6106 ⴚ0.8193
Questions on CEO Surveys #1 #2 #4 #5
Training of hospital staff Equipment Staff for non-drug treatment Formal QA program
#7 Market value to payers #8 Timely, effective treatment #9 Medical liability
a The factor matrix shows the correlations between the original variables and the derived factors. These four factors explain 63.4% of the total variance in the [chief executive officer] (CEO) survey.
tions and pain acuity and increased patient satisfaction. Questions 3 and 6 are related to the fourth factor that concerns reimbursement for some pain management methods as recommended by guidelines. It is noted that Question 10 loads significantly on both Factors 2 and 3, suggesting that reduced postoperative complications not only reflect improved patient outcomes but also have significant legal implications. Question 3 loads significantly on both Factors 1 and 4, suggesting that if reimbursement is a concern for some pain management regimens, expertise may not be available at hospitals to deliver these pain treatments. In summary, the four extracted factors account for 63.4% of the total variance in the CEO survey, which is considered to be a satisfactory solution.
Prediction of Guideline Implementation Results of the logistic regression show how a hospital CEO’s perception of factors affecting guideline implementation was related to the likelihood of the hospital’s adoption of a pain management guideline (Table 5). After controlling for hospital characteristics (i.e., bed size, teaching status, urban/rural location, and sponsorship), three factors were found to be significantly associated with guideline implementation: 1) availability of institutional resources; 2) belief in the favorable impact of guideline usage on patient outcomes; and 3) belief in achieving economic and legal benefits through guideline adoption. Both availability of institutional resources and the belief in improved patient outcomes were strong predictors of guideline implementation. Reimbursement concerns, on the other hand, were not significantly related to the likelihood of adopting a pain management guideline. Logistic regression analysis was used further to identify those factors, as perceived by hospital CEOs, that signifi-
cantly predict the likelihood of complying with key elements of the AHRQ/AHCPR guideline. The results are summarized in Table 6. Availability of institutional reTable 5. Relationship Between the Likelihood of Adopting a Pain Management Guideline and the Hospital CEO’s Perception of Issues Related to Guideline Implementationa 95% Confidence Odds Ratio Interval p-valueb
Independent Variables Hospital CEOs’ perception on issues related to guideline implementationc Institutional resources Economic/legal incentives Patient outcomes Reimbursement concerns Control variables:d Bed size Teaching Urban Voluntary
2.433 1.668 2.526 1.160
(1.417,4.179) (1.025,2.714) (1.323,4.824) (0.731,1.838)
0.0013 0.0394 0.0050 0.5287
1.002 0.551 1.076 1.151
(1.000,1.004) (0.161,1.882) (0.251,4.615) (0.265,5.008)
0.1053 0.3416 0.9215 0.8509
a
Logistic regression model was estimated with the dependent variable defined as 1 if the hospital reported using a nationally published pain management practice guideline and 0 if not. b
p-value: those underscored are significant at the 0.05 level. Hospital CEOs’ perception on issues related to guideline implementation was measured by four composite variables (called factors) that were derived from the 12 questions on the CEO survey (see Table 4). c
d
Teaching status was measured as 1 for teaching hospitals and 0 for nonteaching hospitals. Urban was measured as 1 for urban location and 0 for the rural. Over 95% of the community hospitals in New York State are nonprofit. Sponsorship was measured as 1 for voluntary and 0 for public. J. Clin. Anesth., vol. 13, June 2001
273
Original Contributions
Table 6. Relationship Between the Likelihood of Complying with Key Elements of the AHRQ/AHCPR Acute Pain Management Guideline and the Hospital CEO’s Perception of Issues on Guideline Implementationa Key Elements of the AHRQ/AHCPR Guideline Reported Being Used for Most to All Surgical Inpatients
Hospital CEO’s Perceptions (Independent Factors)
Interdisciplinary Team
Institutional resources Economic/legal incentives Patient outcomes Reimbursement concerns
2.795* (1.662, 1.185 (0.788, 1.266 (0.811, 0.893 (0.592,
4.701) 1.782) 1.976) 1.345)
Pain Control Plan Developed Preoperatively
Documentation of Pain History
Discussion with Patients and Families
Use of Formal Pain Assessment Tool
2.685* (1.640, 4.396) 0.957 (0.642, 1.425) 1.413 (0.920, 2.170) 1.602‡ (1.044, 2.457)
2.449* (1.490,4.022) 0.949 (0.631,1.429) 1.075 (0.710,1.627) 1.502 (0.965,2.337)
2.065† (1.275, 3.345) 1.036 (0.690, 1.556) 1.387 (0.889, 2.163) 1.556‡ (1.005, 2.410)
2.375* (1.500,3.760) 1.232 (0.829,1.833) 1.176 (0.779,1.775) 1.263 (0.843,1.894)
a
Logistic regression model was estimated for each key element, with the dependent variable defined as 1 if the hospital reported implementing this key element for most to all surgical inpatients, and 0 if not. Odds ratio and 95% confidence intervals (CIs)are presented. Each model has been controlled for hospital characteristics including bed size, teaching status, urban/rural location, and sponsorship. AHRQ/AHCPR ⫽ Agency for Healthcare Research and Quality/Agency for Healthcare Policy and Research, CEO⫽chief executive officer. *p ⬍ 0.001. p ⬍ 0.01. p ⬍ 0.05.
† ‡
sources was found to be a significant predictor for compliance with five key elements of the AHRQ/AHCPR guideline: use of an interdisciplinary team, development of a pain control plan preoperatively, documentation of the patient’s pain history, discussion with patients and families about pain management, and use of a formal pain assessment tool. The results also indicate that when reimbursement was considered an issue, a hospital was more likely to have an individualized pain control plan developed preoperatively and include patients and families in the discussion. Given the importance of resource availability to guideline implementation, additional analysis was conducted to examine the association of each resource aspect with use of guidelines and compliance of five key elements of the AHRQ/AHCPR guideline (Table 7). First, hospitals were stratified into two groups based on their CEOs’ responses to each of five questions concerning resource availability; that is, whether the hospital CEO mostly/strongly agreed or mostly/strongly disagreed with each resource statement. Then a comparison was made in the percentage of hospitals reporting using a guideline between the two groups of hospitals (i.e., “agree” vs. “disagree” on each resource aspect). Clinical directors reported using a pain management guideline significantly more often in hospitals where staff are trained and knowledgeable in pain management, particularly nondrug intervention. Similar findings were shown for compliance with key elements of the AHRQ/ AHCPR guideline. In particular, existence of a formal QA program to monitor pain management was significantly associated with use of a pain management guideline and compliance with several key elements of the AHRQ/ AHCPR guideline. Availability of enough specialists for epidural catheters was significantly associated with hospitals complying with two key elements: development of a 274
J. Clin. Anesth., vol. 13, June 2001
pain control plan preoperatively and inclusion of the patients and families in the formulation of the plan.
Discussion Based on the survey results, a majority of hospital CEOs agreed that their clinical staff were generally trained and knowledgeable in pain management, and that they had the equipment necessary to deliver the recommended interventions. Most of the surveyed CEOs also believed that implementing a pain management practice guideline was beneficial in terms of lowered cost, improved quality of care, and reduced risk of medical liability. Nevertheless, only 27% of the surveyed hospitals reported using a nationally published clinical practice guideline. Less than 50% of the respondents said that their current pain management programs contained those key elements as recommended by the AHRQ/ AHCPR guideline such as use of an interdisciplinary approach and an individualized proactive pain control plan. There also appeared to be a lack of staff trained in delivering nondrug interventions and an under-utilization of nonpharmacological therapies to control pain. Further analysis of the survey results using factor analysis and logistic regression suggested that three major factors were significantly associated with the implementation of pain management practice guidelines. The first factor was availability of institutional resources, particularly the required personnel for pain management and the existence of a formal QA program or other institutional program to monitor pain management. These findings confirm some of the concerns raised by Carr and colleagues regarding local resource shortfalls for implementing pain management practice guidelines.22 The second impetus for hospitals to adopt pain management guidelines was to improve patient outcomes including reduced
Acute pain management guidelines: Jiang et al.
Table 7. Association of Guideline Usage with Hospital CEOs’ Perception of Resource Availability
Resource Availability (based on the CEOs’ response to each of the following statements) Hospital staff is trained and knowledgeable in pain management –Agree –Disagree Hospital has the equipment to deliver various pain treatment –Agree –Disagree Hospital has enough specialists with expertise to do epidural catheters –Agree –Disagree Hospital has enough trained staff to deliver non-drug interventions –Agree –Disagree Hospital has a formal QA program to monitor pain management –Agree –Disagree
Key Elements of the AHRQ/AHCPR Guideline Reported Being Used for Most to All Surgical Inpatients (% who reported implementing for most to all inpatient surgical patients)
Use of a Pain Management Guideline (% Interdisciplinary reporting “Yes”) Team
Developing Pain Control Plan
Documentation of Pain History
Discussion With Patients/ Families
Use of Pain Assessment Tool
33.3% 11.8% (p ⫽ 0.016)
52.1% 17.1% (p ⫽ 0.000)
45.8% 28.6% (p ⫽ 0.076)
40.6% 14.3% (p ⫽ 0.005)
40.6% 14.3% (p ⫽ 0.005)
55.2% 34.3% (p ⫽ 0.034)
29.8% 7.7% (p ⫽ 0.091)
44.9% 23.1% (p ⫽ 0.131)
43.2% 23.1% (p ⫽ 0.161)
33.1% 38.5% (p ⫽ 0.695)
33.1% 38.5% (p ⫽ 0.695)
53.4% 15.4% (p ⫽ 0.009)
29.1% 17.6% (p ⫽ 0.326)
45.1% 27.8% (p ⫽ 0.167)
45.1% 16.7% (p ⫽ 0.023)
35.4% 22.2% (p ⫽ 0.272)
37.2% 11.1% (p ⫽ 0.030)
52.2% 33.3% (p ⫽ 0.137)
41.5% 20.9% (p ⫽ 0.015)
55.8% 36.4% (p ⫽ 0.035)
60.5% 31.8% (p ⫽ 0.002)
60.5% 20.5% (p ⫽ 0.000)
55.8% 22.7% (p ⫽ 0.000)
55.8% 46.6% (p ⫽ 0.321)
33.7% 14.6% (p ⫽ 0.024)
47.2% 33.3% (p ⫽ 0.135)
48.3% 26.2% (p ⫽ 0.016)
40.4% 19.0% (p ⫽ 0.015)
38.2% 23.8% (p ⫽ 0.104)
58.4% 31.0% (p ⫽ 0.003)
p-value underlined denotes that it is significant at the 0.05 level. CEO ⫽ chief executive officer, AHRQ/AHCPR ⫽ Agency for Healthcare Research and Quality/Agency for Healthcare Policy and Research, QA ⫽ quality assurance.
pain severity and enhanced patient satisfaction. The third driving force for guideline implementation related to economic and legal incentives such as enhanced market value to payers and reduced medical liability. The potential of reducing medical liability through guideline implementation derives from the premise that practice guidelines are built on evidence-based medicine and, therefore, can justify or assist physicians in their clinical decision making.26 Apparently, the first major factor determining the implementation of pain management practice guidelines was a hospital’s internal resources, while the second and third factors reflected individual drive toward clinical excellence and expectations from a variety of external audiences, including payers, consumers, and regulatory entities. Although 65% of the responding CEOs considered reimbursement for some pain management methods an issue for implementing pain management practice guidelines, this factor was not found to be significantly associated with the adoption of guidelines. One possible explanation is that when considered together with all other
factors in the logistic regression analysis, the influence of this variable became relatively small. Concerns for reimbursement of certain recommended interventions did not seem to impede the use of pain management practice guidelines. Nevertheless, the results of this study also indicate that when reimbursement was considered an issue, hospitals were less likely to have expertise available to deliver some specialized pain management regimens. Facilitating consistent payment criteria is therefore important to support appropriate use of all available pain treatment methods. Several limitations of this study need to be addressed. First, our survey sample includes only hospitals located in New York State, which may limit the generalizability of our findings to hospitals in other states. Conversely, New York hospitals represent a wide variety of providers, including major teaching hospitals located in large metropolitan areas such as New York City and small community hospitals serving rural or mountainous communities in the upstate area. This situation allowed us to observe certain J. Clin. Anesth., vol. 13, June 2001
275
Original Contributions
variations across different hospitals. Second, questions on the CEO survey that related to availability of institutional resources were only proxy measures based on the CEO’s perception. Perhaps it would have been more accurate to measure the size and composition of the pain management staff, the availability of equipment, and other organizational attributes. Third, the CEO survey may not have included all factors affecting the implementation of pain management guidelines. These factors could include the existence of regulatory obstacles that may inhibit appropriate pain management, and the influence of accreditation bodies such as the JCAHO. In conclusion, the findings of this study contribute to the understanding of factors influencing the implementation of pain management practice guidelines in hospital settings. The external environment can generate a variety of incentives for using pain management practice guidelines, but internal resource shortfalls may significantly inhibit implementation. In light of the fact that pain management has recently been incorporated in the JCAHO accreditation process, this information is both timely and important for both clinicians and health care executives. Many hospitals may consider implementing pain management practice guidelines as part of their efforts to comply with the JCAHO standards. To ensure the successful implementation of such guidelines, it is important for hospitals to carefully assess their internal resources, including the staff’s knowledge and skills in pain management, and to establish an institutional approach that results in a measurable improvement in outcomes. Implementing practice guidelines can contribute to reduced pain severity and enhanced patient satisfaction. Strategies are still needed for developing other effective mechanisms to support and encourage desired practice. Ultimately, real improvements in diagnosis and treatment of pain will require a coordinated effort of public policy, providers, patients, and payers.
Acknowledgments The authors wish to acknowledge Onita Munshi, M.S., MBA, a former researcher of HANYS, for her indispensable assistance with survey construction and administration, data collection, and analysis at the early stage of this project.
References 1. Pain terms: a list with definitions and notes on usage. Recommened by the IASP. Pain 1979;6:249. 2. Abbott FV, Gray-Donald K, Sewitch MJ, Johnston CC, Edgar L, Jeans ME: The prevalence of pain in hospitalized patients and resolution over six months. Pain 1992;50:15–28. 3. Donovan M, Dillon P, McGuire L: Incidence and characteristics of pain in a sample of medical-surgical inpatients. Pain 1987;30: 69 –78. 4. Lynch EP, Lazor MA, Gellis JE, Orav J, Goldman L, Marcantonio ER: Patient experience of pain after elective noncardiac surgery. Anesth Analg 1997;85:117–123. 5. Oden R: Acute postoperative pain: incidence, severity, and the etiology of inadequate treatment. Anesth Clin N Am 1989;7:1–15. 6. Agency for Healthcare Research and Quality, formerly Agency for Health Care Policy and Research: Acute pain management: operative or
276
J. Clin. Anesth., vol. 13, June 2001
7.
8.
9.
10.
11. 12. 13.
14. 15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
medical procedures and trauma. AHRQ Publication No. 92–0032. Rockville, MD: US Department of Health and Human Services, 1992. Am Pain Society: Am Pain Society quality assurance standards for relief of acute pain and cancer pain. In: Bond MR, Charlton JE, Woolf CJ (eds): Proceedings of the VIth World Congress on Pain. New York: Elsevier Science Publishers, 1991. Practice guidelines for acute pain management in the perioperative setting. A report by the Am Society of Anesthesiologist task force on pain management, acute pain section. Anesthesiology 1995;82:1071– 81. Bach DM: Implementation of the Agency for Health Care Policy and Research postoperative pain management guideline. Nurs Clin North Am 1995;30:515–27. Bookbinder M, Coyle N, Kiss M, et al: Implementing national standards for cancer pain management: program model and evaluation. J Pain Symptom Manage 1996;12:334 –347. Campese C: Development and implementation of a pain management program. AORN J 1996;64:937– 40. Knowles R: Standardization of pain management in the postanesthesia care unit. J Perianesths Nurs 1996;11:390 – 8. Davis DA, Taylor-Vaisey A: Translating guidelines into practice. A systematic review of theoretic concepts, practical experience and research evidence in the adoption of clinical practice guidelines. CMAJ 1997;157:408 –16. Rogers EM: Lessons for guidelines from the diffusion of innovations. Jt Comm J Qual Improv 1995;21:324 – 8. James PA, Cowan TM, Graham RP, Majeroni BA: Family physicians’ attitudes about and use of clinical practice guidelines. J Fam Pract 1997;45:341–7. Lawler FH, Viviani N: Patients and physician perspectives regarding treatment of diabetes: Compliance with practice guidelines. J Fam Pract 1997;44:369 –73. Weingarten S, Stone E, Hayward R, et al: The adoption of preventive care practice guidelines by primary care physicians: do actions match intentions? Gen Intern Med 1995;10:138 – 44. Reis SE, Holubkov R, Zell KA, Edmundowicz D, Shapiro AH, Feldman AM: Unstable angina: specialty-related disparities in implementation of practice guidelines. Clin Cardiol 1998;21:207–10. Ely JW, Goerdt CJ, Bergus GR, West CP, Dawson JD, Doebbeling BN: The effect of physician characteristics on compliance with adult preventive care guidelines. Fam Med 1998;30:34 –9. Goodpastor WA, Montoya ID: Motivating physician behavior change: Social influence versus financial contingencies. Int J Health Care Qual Assur 1996;9:4 –9. Petrazzuoli M, Rothe MJ, Grin-Jorgensen C, Ramsey WH, GrantKels JM: Monitoring patients taking methotrexate for hepatotoxicity. Does the standard of care match published guidelines? J Am Acad Dermatol 1994;31:969 –77. Carr D, Jacox A, Chapman C, et al.: Acute pain management: guideline technical report (No.1). Rockville, MD: Agency for Health Care Policy and Research, Public Health Service, U.S. Department of Health and Human Services; 1995. Hair JF Jr., Anderson RE, Tatham RL, Black WC: Multivariate Data Analysis with Readings. New York, NY: Macmillan Publishing Company, 1992. Ware JE Jr, Kosinski M, Gandek B, et al: The factor structure of the SF-36 Health Survey in 10 countries: results from the IQOLA Project. International Quality of Life Assessment. J Clin Epidemiol 1998;51:1159 – 65. Zaslavsky AM, Beaulieu ND, Landon BE, Cleary PD: Dimensions of consumer-assessed quality of Medicare managed-care health plans. Med Care 2000;38:162–74. Merritt TA, Palmer D, Bergman DA, Shiono PH: Clinical practice guidelines in pediatric and newborn medicine: implications for their use in practice. Pediatrics 1997;99:100 –14.