Abstract
OBJECTIVES
To assess physician awareness and reported use of medical guidelines for community-acquired pneumonia (CAP), and to identify factors associated with variations in awareness and use of these guidelines.
DESIGN
A questionnaire was administered during the preintervention phase of a randomized clinical trial of a pneumonia guideline implementation strategy.
PARTICIPANTS
Three hundred and fifty-two physicians who managed CAP patients at 7 Pittsburgh, PA hospitals completed the questionnaire. Physician and practice setting characteristics, and physician awareness and reported use of national American Thoracic Society (ATS) and local (hospital-developed) guidelines for CAP were assessed.
RESULTS
Overall, 48% reported being influenced by ATS guidelines and 20% reported using these guidelines; 48% were uncertain whether a local pneumonia guideline existed. Only 28% of physicians who knew a local guideline existed reported frequently using the guideline. Use of national ATS guidelines was independently associated with practice as an infectious disease or pulmonary medicine specialist, nonpatient care-related professional activities, and intellect personality score. Use of local guidelines was independently negatively associated with practice as an infectious disease or pulmonary medicine specialist, and positively associated with positive attitudes toward practice guidelines.
CONCLUSIONS
Results indicate low levels of awareness and use of guidelines for the management of CAP. Key indicators (e.g., medical specialty, fewer clinical duties, and positive attitudes about guidelines) were associated with greater use of national and local guidelines. If replicated with data on actual physician management practices, more effective guideline implementation strategies will be necessary to encourage compliance with practice guidelines for the management of CAP.
Keywords: community-acquired pneumonia, medical practice guidelines
During the past two decades, widespread variation in the treatment of specific medical conditions has been documented. This variation is reflected in differences in hospital admissions rates,1 in the use of procedures,2 and in length of hospital stay (LOS).3,4 It has also been an important contributing factor to the rise of utilization management5 and the development of medical practice guidelines as tools to improve the quality and cost-efficacy of health care.6–10
Previous investigations suggest that there is large variation in the success of guideline implementation.11–13 In examining this variation, most previous guideline implementation studies have controlled for the effects of physician and practice characteristics on guideline use rather than examining their effects systematically—despite the fact that such characteristics have been identified as key components in effective adoption of guidelines.14,15 Our goal in the current investigation was to focus on two key elements of guideline adherence15—awareness and adoption of guidelines—and physician and practice setting characteristics associated with these elements in the context of community-acquired pneumonia (CAP).
We focused on CAP as a model for several reasons. First, it is a major contributor to morbidity and mortality in the United States, causing 4 million episodes of illness and more than 1 million hospitalizations annually.16 Second, the considerable geographic variation in hospital admission rates and processes of care for CAP2,3,17–20 have led to the development of numerous national and local medical practice guidelines for CAP—for example, the American Thoracic Society (ATS) in 1993 and 2001,7,8 the Infectious Disease Society of America (IDSA) in 1998 and 2000,9,10 and numerous healthcare organizations on a local level. Finally, little is known about physician awareness and use of guidelines for CAP and factors associated with guideline use. Thus, the goals of the current study were (1) to assess physician awareness and use of nationally developed ATS and locally developed hospital guidelines for CAP, and (2) to identify hospital and physician characteristics associated with awareness and reported use of these guidelines.
Although more than 80% of patients with CAP are treated by primary care physicians (e.g., internal medicine, family medicine), the two national guidelines were developed by pulmonary medicine and infectious disease specialists. Thus, we were particularly interested in comparing the awareness and use of CAP guidelines among generalist physicians and medical specialists. In addition to characteristics of the practice setting, we also expected that other physician demographic, professional practice (e.g., workload, continuing medical education), and personality characteristics (e.g., self-rated “intellectualism” and “agreeableness”), might be related to reported awareness and use of guidelines.
METHODS
During the preintervention phase of a randomized clinical trial of a guideline dissemination strategy to reduce duration of parenteral antibiotic therapy and length of hospital stay for patients with CAP,21,22 a questionnaire examining physicians’ awareness and use of national and local guidelines for the management of patients hospitalized with CAP was administered.
Study Sites and Participants
Seven hospitals in Pittsburgh, PA (1 university, 3 community teaching, and 3 community nonteaching) participated in the study. Six hundred and twenty-one physicians with staff privileges at one or more of the 7 hospitals were identified as meeting one of the following inclusion criteria: (1) practice in a primary care medical specialty; (2) practice in an internal medicine subspecialty of pulmonary medicine or infectious diseases; or (3) practice in another internal medicine subspecialty and documentation of two or more pneumonia admissions at their primary study site hospital in 1996 based on claims data (DRGs 79, 80, 89, 90).
Questionnaire and Measures
A questionnaire (available from the authors upon request) was developed to assess physician demographic, professional, and psychosocial characteristics. Information about hospital characteristics was gathered from each hospital. Finally, study outcomes about physician awareness and use of ATS and local guidelines for CAP were assessed.
Physician Characteristics
Physician demographic characteristics included gender, age, and years since medical school graduation. Physician professional characteristics included medical specialty; hours/week in direct outpatient care, direct inpatient care, total patient care, total nonpatient activities (e.g., teaching, administration, research), and total job-related activities; number of CAP inpatients treated per year; percent income from salary; frequency of journal reading; hours/month spent on quality assurance (QA) activities; hours/month spent on continuing medical education (CME) activities; satisfaction with practice autonomy; and general career satisfaction. Medical specialties were characterized as (1) generalists (i.e., general internal medicine, family practice, geriatrics, community medicine), (2) pulmonary medicine and infectious diseases specialists, and (3) other specialists (e.g., cardiology, gastroenterology, rheumatology).
Physician psychosocial characteristics included 3 risk avoidance items from the Jackson Personality Inventory,23 (e.g., “I rarely take risks when there is another alternative”). Response categories ranged from 0 (strongly disagree) to 4 (strongly agree). Items were averaged to form a scale ranging from 0 (low risk avoidance) to 4 (high risk avoidance). Two personality traits, “agreeableness” and “intellectualism,” from the Goldberg Personality Scale24 were assessed with 10 items each. The scale consists of sets of adjectives (e.g., cooperative, harsh for “agreeableness”, and introspective, uncreative for “intellectualism”) rated on a 5-point scale ranging from 0 (not at all accurate) to 4 (extremely accurate). Items were averaged to form two composite scales, ranging from 0 (low on trait) to 4 (high on trait). Finally, general attitudes about guidelines were assessed with the 9-item Tunis General Guideline Experience and Attitudes scale.25 The Tunis scale includes items asking whether guidelines are “good educational tools,” and “a challenge to physician autonomy.” Responses were rated on a 5-point scale ranging from 0 (strongly disagree) to 4 (strongly agree). Items were summed to form a composite scale ranging from 0 (negative attitudes) to 40 (positive attitudes).
Hospital Characteristics
Hospital characteristics included size (small, <200 beds, n = 1; medium, 200 to 500 beds, n = 5; large, >500 beds, n = 1), teaching versus nonteaching, and university-affiliated versus nonuniversity-affiliated. It was also known for each study site, whether or not local pneumonia guidelines existed at the time the survey was conducted.
Dependent Variables
Guideline awareness and reported use were assessed with two items concerning ATS guidelines and two items concerning local guidelines. Questions were based on the ATS guidelines because the IDSA pneumonia guidelines had not been published or were broadly disseminated at the time our survey was designed. Questions about ATS guidelines were “How familiar are you with the American Thoracic Society guidelines for the management of community-acquired pneumonia?” (1 = not at all familiar, 2 = I’ve seen them, 3 = I’ve read them, 4 = I use them), and “How influential have the American Thoracic Society guidelines been on your management of patients with community-acquired pneumonia?” (1 = not influential to 4 = very influential). Questions about locally developed guidelines were “Is there a medical practice guideline (or critical pathway, care map, or algorithm) for the management of community-acquired pneumonia at your hospital?” (yes, no), and if yes, “Please estimate the percentage of community-acquired pneumonia patients you treated over the past 12 months using the hospital's guideline.” Original response categories for this variable (≤25%, 26% to 50%, 51% to 75%, >75%) were dichotomized (≤50%, >50%).
Questionnaire Administration
The questionnaire was pilot-tested with 24 physicians. Questionnaires were mailed to the 621 eligible physicians in January 1998. A second questionnaire was mailed to nonresponders 2 weeks later, and telephone calls were placed to nonresponders 10 days after the second mailing. A local opinion leader serving as the local study investigator at each site personally distributed a third questionnaire to those who did not respond to the first two mailings and telephone contact. All questionnaire data were collected between January 13, 1998 and June 11, 1998.
Analysis
The relationship of study variables and guideline use was examined with χ2 tests for categorical variables and nonparametric, two-sample Mann-Whitney tests for continuous variables. All tests were two-tailed, and an alpha of P≤ .05 was considered statistically significant. Variables that had bivariate associations with their respective outcomes of alpha ≤ .10 were entered stepwise into a multivariate logistic regression analysis. The following variables were not significantly related to use of ATS or local guidelines in bivariate analyses, and were not included in the summary of findings or in Tables 1–4: gender, age, years since graduation from medical school, university hospital affiliation, percent of income from salary, total patient care hours per week, CME hours per month, and personality trait “agreeable.”
Table 1.
Factors Related to Self-Reported Use of American Thoracic Society Guidelines for the Management of Community-acquired Pneumonia
| Hospital and Physician Characteristics* | Total N | Report Not Using ATS Guidelines† | Report Using ATS Guidelines† | Bivariate P Value | Multiple Stepwise Odds Ratio (95% CI)‡ |
|---|---|---|---|---|---|
| Hospital characteristics | |||||
| Hospital size, % | .34 | ||||
| Small (<200 beds) | 11 | 82 | 18 | ||
| Medium (200∠500 beds) | 181 | 76 | 24 | ||
| Large (>500 beds) | 72 | 68 | 32 | ||
| Teaching status, % | .27 | ||||
| Teaching | 180 | 79 | 21 | ||
| Nonteaching | 84 | 72 | 28 | ||
| Physician characteristics | |||||
| Specialty, % | <.001 | ||||
| Generalist | 210 | 77 | 23 | 1.00 | |
| Pulmonary or ID specialist | 28 | 43 | 57 | 4.51*** (1.71 to 11.90) | |
| Other specialist | 26 | 85 | 15 | 0.80* (0.25 to 2.58) | |
| Direct inpatient care, h/wk ± SD | 259 | 14.4 ± 13.4 | 18.0 ± 16.8 | .24 | |
| Direct outpatient care, h/wk ± SD | 260 | 30.2 ± 15.4 | 23.4 ± 13.0 | .001 | – |
| Total nonpatient activities, h/wk ± SD | 235 | 10.6 ± 12.0 | 17.9 ± 15.5 | <.001 | 1.02** (1.00 to 1.05) |
| Percent of total time on inpatient care ± SD | 248 | 25.1 ± 20.9 | 29.5 ± 24.5 | .32 | |
| Total work, h/wk ± SD | 250 | 56.1 ± 13.1 | 57.8 ± 12.6 | .44 | |
| Number of CAP inpatients per year ± SD | 259 | 22.4 ± 31.6 | 27.0 ± 28.3 | .07 | – |
| Journal reading, % | .02 | – | |||
| Do not read or skim articles | 125 | 81 | 19 | ||
| Read ≥1 article in most issues | 137 | 69 | 31 | ||
| Quality assurance hours, % | .26 | ||||
| <1 h per month | 155 | 77 | 23 | ||
| ≥1 h per month | 109 | 71 | 29 | ||
| Intellect personality score (0–4 = high)§± SD | 232 | 2.82 ± 0.50 | 2.99 ± 0.55 | .01 | 2.18** (1.07 to 4.40) |
| Attitudes about guidelines (0–40 = pos)″ ± SD | 258 | 16.9 ± 4.7 | 18.6 ± 4.8 | .02 | – |
| Risk avoidance (0–4 = high) ± SD | 257 | 3.67 ± 0.82 | 3.76 ± 0.92 | .45 |
Gender, age, years since graduation from medical school, university hospital affiliation, percent of income from salary, total patient care hours per week, continuing medical education hours per month, and personality trait “agreeable” were not significantly related to awareness and use of guidelines or any of the other guideline outcomes presented in Tables 1–4.
Percentages are calculated within each row.
P value < .10 but ≥ .05,
P value < .05 but ≥ 0.01,
P value < .01.
The Goldberg Personality Scale is self-rated on a 5-point response scale ranging from 0 (not at all accurate) to 4 (extremely accurate). Items were averaged to form a composite scale ranging from 0 (low personality trait intellectualism) to 4 (high personality trait intellectualism).
Attitudes about guidelines in general were assessed with the 9-item Tunis General Guideline Experience and Attitudes scale. Items were averaged to form a composite scale ranging from 0 (negative attitudes) to 4 (positive attitudes).
CAP, community-acquired pneumonia.
Table 4.
Factors Related to Self-reported Use of Local Community-acquired Pneumonia Guidelines
| Hospital and Physician Characteristics* | Total N | Use Local Guidelines for ≤50% of CAP Patients† | Use Local Guidelines for >50% of CAP Patients† | Bivariate P Value | Multiple Stepwise Odds Ratio (95% CI)‡ |
|---|---|---|---|---|---|
| Hospital characteristics | |||||
| Hospital size, % | .28 | ||||
| Small (<200 beds) | 1 | 0 | 100 | ||
| Medium (200−500 beds) | 82 | 70 | 30 | ||
| Large (>500 beds) | 27 | 74 | 26 | ||
| Teaching status, % | .68 | ||||
| Teaching | 77 | 69 | 31 | ||
| Nonteaching | 33 | 73 | 27 | ||
| Physician characteristics | |||||
| Specialty, % | .02 | ||||
| Generalist | 85 | 64 | 36 | 1.00 | |
| Pulmonary or ID specialist | 13 | 92 | 8 | 0.10** (0.01 to 0.89) | |
| Other specialist | 12 | 92 | 8 | 0.11* (0.01 to 1.07) | |
| Direct inpatient care, h/wk ± SD | 106 | 17.4 ± 14.7 | 16.5 ± 14.7 | .93 | |
| Direct outpatient care, h/wk ± SD | 107 | 27.9 ± 16.6 | 31.8 ± 13.2 | .18 | |
| Total nonpatient activities, h/wk ± SD | 97 | 12.5 ± 14.7 | 13.5 ± 12.9 | .43 | |
| Percent of total time on inpatient care ± SD | 101 | 29.9 ± 23.4 | 26.7 ± 20.0 | .73 | |
| Total work, h/wk ± SD | 101 | 58.3 ± 12.3 | 58.3 ± 13.2 | .95 | |
| Number of CAP inpatients per year ± SD | 108 | 24.7 ± 29.7 | 26.1 ± 33.1 | .53 | |
| Journal reading, % | .59 | ||||
| Do not read or skim articles | 55 | 73 | 27 | ||
| Read ≥1 article in most issues | 52 | 68 | 32 | ||
| Quality assurance hours, % | .67 | ||||
| <1 h per month | 70 | 71 | 29 | ||
| ≥1 h per month | 40 | 68 | 32 | ||
| Intellect personality score (0–4 = high) ± SD | 96 | 2.83 ± 0.52 | 3.02 ± 0.46 | .09 | – |
| Attitudes about guidelines (0–40 = pos) ± SD | 108 | 16.9 ± 4.7 | 20.5 ± 3.6 | <.001 | 1.29*** (1.09 to 1.47) |
| Risk avoidance (0–4 = high) ± SD | 106 | 3.78 ± 0.80 | 3.85 ± 0.76 | .67 |
Gender, age, years since graduation from medical school, university hospital affiliation, percent of income from salary, total patient care (h/wk), continuing medical education hours per month, and personality trait “agreeable” were not significantly related to awareness and use of guidelines or any of the other guideline outcomes presented in Tables 1–4.
Percentages are calculated within each row.
P value < .10 but ≥ .05,
P value < .05 but ≥ .01,
P value < .01.
CAP, community-acquired pneumonia.
RESULTS
Response Rates and Respondent Characteristics
Three hundred and forty-five physicians completed the questionnaire (56% response rate). The majority of respondents were generalists (79%), white (78%), and male (81%). The mean age of respondents was 47.6 years (SD = 11.6), and the mean years since completing medical school was 20.4 (SD = 12.2). The majority of respondents practiced at medium-sized hospitals (70%) versus small (6%) and large (24), teaching hospitals (65%), and nonuniversity hospitals (75%). Response rates differed by hospital site (range 42%το 68%), hospital size (range 52%το 68%), university affiliation (52% vs 68%), and physician medical specialty (range 42% to 60%).
Self-Reported Use of National ATS Guidelines
Twenty-one percent of respondents (range 11% to 42% by site) reported being not at all familiar with ATS guidelines, 29% (18%το 42% by site) reported having seen the guidelines, 30% (16% to 39% by site) reported having read the guidelines, and 20% (8% to 28% by site) reported using the guidelines. This variable was dichotomized (use guidelines vs do not use guidelines) for subsequent analyses; the 75 physicians who were not familiar with ATS guidelines were excluded from the analyses presented in Tables 1 and 2.
Table 2.
Factors Related to the Influence of American Thoracic Society Guidelines on the Management of Community-acquired Pneumonia
| Hospital and Physician Characteristics* | Total N | Not/Slightly Influenced† | Moderately/Very Influenced† | Bivariate P Value | Multiple Stepwise Odds Ratio (95% CI)‡ |
|---|---|---|---|---|---|
| Hospital characteristics | |||||
| Hospital size, % | .06 | – | |||
| Small (<200 beds) | 11 | 73 | 27 | ||
| Medium (200−500 beds) | 181 | 36 | 64 | ||
| Large (>500 beds) | 70 | 40 | 60 | ||
| Teaching status, % | .53 | ||||
| Teaching | 178 | 42 | 58 | ||
| Nonteaching | 84 | 38 | 62 | ||
| Physician characteristics | |||||
| Specialty, % | .10 | – | |||
| Generalist | 208 | 42 | 58 | ||
| Pulmonary or ID specialist | 28 | 21 | 79 | ||
| Other specialist, % | 26 | 35 | 65 | ||
| Direct inpatient care, h/wk ± SD | 257 | 13.4 ± 13.0 | 16.6 ± 15.2 | .09 | – |
| Direct outpatient care, h/wk ± SD | 258 | 30.8 ± 15.4 | 27.0 ± 14.8 | .02 | 0.98** (0.96 to 1.00) |
| Total nonpatient activities, h/wk ± SD | 233 | 11.2 ± 13.2 | 13.6 ± 13.6 | .06 | – |
| Percent of total time on inpatient care | 246 | 23.2 ± 19.3 | 28.3 ± 23.3 | .10 | – |
| Total work, h/wk ± SD | 248 | 56.4 ± 12.5 | 56.9 ± 12.9 | .75 | |
| Number of CAP inpatients per year ± SD | 257 | 21.6 ± 30.8 | 25.1 ± 30.9 | .20 | |
| Journal reading | .22 | ||||
| Do not read or skim articles | 124 | 43 | 57 | ||
| Read ≥1 article in most issues | 136 | 35 | 65 | ||
| Quality assurance hours, % | .10 | – | |||
| <1 h per month | 153 | 43 | 57 | ||
| ≥1 h per month | 109 | 33 | 67 | ||
| Intellect personality score (0–4 = high) ± SD | 230 | 2.79 ± 0.52 | 2.91 ± 0.51 | .19 | |
| Attitudes about guidelines (0–40 = pos) ± SD | 256 | 16.5 ± 5.1 | 17.8 ± 4.5 | .08 | 1.05* (0.99 to 1.11) |
| Risk avoidance (0–4 = high) ± SD | 255 | 3.66 ± 0.79 | 3.72 ± 0.88 | .46 |
Gender, age, years since graduation from medical school, university hospital affiliation, percent of income from salary, total patient care h/wk, continuing medical education hours per month, and personality trait “agreeable” were not significantly related to awareness and use of guidelines or any of the other guideline outcomes presented in Tables 1–4.
Percentages are calculated within each row.
P value < .10 but ≥ .05,
P value < .05 but ≥ .01,
P value < .01.
CAP, community-acquired pneumonia.
No hospital characteristics were associated with use of ATS guidelines. Physicians who reported using ATS guidelines were more likely to be pulmonary and ID specialists (57% vs 23% generalists and 15% other specialists, P < .001), to spend fewer hours/week in direct outpatient care (23.4 ± 13.0 vs 30.2 ± 15.4, P = .001), and to spend more hours/week in nonpatient-related activities (17.9 ± 15.5 vs 10.6 ± 12.0, P < .001). In addition, physicians who reported using ATS guidelines read journals more frequently (31% vs 19%, P < .05), were higher on personality trait “intellect” (2.99 ± 0.55 vs 2.82 ± 0.50, P = .01), and had more positive attitudes about guidelines in general (18.6 ± 4.8 vs 16.9 ± 4.7, P < .05). Multivariate analyses indicated that three variables—practicing as a pulmonary or ID specialist (OR = 4.51; 95% CI, 1.71 to 11.90), spending more time in nonpatient-related activities (OR = 1.02; 95% CI, 1.00 to 1.05), and scoring higher on the personality “intellect” score (OR = 2.18; 95% CI, 1.07 to 4.40)—were independently associated with reported use of ATS guidelines.
Influence of ATS Guidelines
Of physicians familiar with ATS guidelines, 48% reported that the guidelines were moderately/very influential on their management of patients. As indicated in Table 2, physicians practicing at large and medium-sized hospitals (60% and 64% vs 27%, P = .06) were marginally significantly more likely to report being influenced by ATS guidelines. In addition, physicians more likely to report being influenced by ATS spent fewer hours/week in direct outpatient care (27.0 ± 14.8 vs 30.8 ± 15.4, P < .05), and spent more hours/week in nonpatient activities (13.6 ± 13.6 vs 11.2 ± 13.2, P = .06). In multivariate analyses, being influenced by ATS guidelines was associated with fewer hours/week in direct outpatient care (OR = 0.98; 95% CI, 0.96 to 1.00) only marginally significant, and more positive attitudes about guidelines (OR = 1.05; 95% CI, 0.99 to 1.11).
Knowledge Accuracy About Local Hospital Pneumonia Guidelines
Six of the 7 study hospitals had local CAP guidelines. Less than half of respondents (42%) were accurate about whether their primary hospital had a practice guideline for CAP. For the 290 respondents from the 6 hospitals with a CAP guideline, 41% reported that no local guideline existed; for the 48 respondents from the hospital with no guideline, 14% reported that one did exist. Physicians from nonteaching hospitals were more accurate in their knowledge about the existence of a local guideline (50% vs 39%, P < .05). Physicians with accurate knowledge about local guidelines spent more hours/week in direct inpatient care (16.5 ± 14.5 vs 14.4 ± 14.3, P < .05), and saw more CAP inpatients per year (25.3 ± 32.3 vs 18.4 ± 25.4, P < .05). Finally, physicians with accurate knowledge were marginally more likely to spend more time on inpatient care (27.8 ± 21.7 vs 25.0 ± 22.6, P = .06), spend more total hours/week working (57.9 ± 13.2 vs 55.2 ± 12.7, P = .06), and were more risk avoidant (3.81 ± 0.80 vs 3.62 ± 0.85, P = .06). In multivariate analyses, accurate knowledge about local guidelines was associated with more total work hours/week (OR = 1.01; 95% CI, 1.00 to 1.02), more positive attitudes about CAP guidelines (OR = 1.06; 95% CI, 1.01 to 1.11), and risk avoidance (OR = 1.29; 95% CI, 0.96 to 1.72).
Use of Local Hospital Pneumonia Guidelines
We next examined the determinants of local guideline use among the 117 physicians who accurately identified their hospital as having a local CAP guideline (see Table 4). Only 30% of respondents from hospitals with a local guideline reported that they used the guideline more than half of the time in management of their CAP patients (range 0% to 58% by site for sites with more than one physician). Physicians who reported using local CAP guidelines more than half of the time were more likely to be generalists (36% vs 8% pulmonary/ID and 8% other specialists, P < .05), and to report more positive attitudes about guidelines (20.5 ± 3.6 vs 16.9 ± 4.7, P < .001). In multivariate analyses, use of local CAP guidelines was associated with practicing as a generalist (OR = 0.10 and 0.11 specialist groups vs generalists), and more positive attitudes about guidelines (OR = 1.29; 95% CI, 1.09 to 1.47).
DISCUSSION
Our goals in the current study were to assess physician awareness and use of national and local CAP guidelines, and to identify predictors of awareness and use. Only approximately half of physicians surveyed reported being influenced by ATS guidelines, and only 20% reported actually using ATS guidelines. Similarly, only half of physicians were correct about the existence of a local hospital CAP, and among these, only 28% reported using the local guideline in managing more than half of their CAP patients. Thus, for both national and local guidelines, simple awareness of the guideline was not sufficient to promote actual use of the guideline in caring for patients. The low rates of CAP guideline use is consistent with other studies of guideline compliance,26,27 and suggests that robust interventions aimed not only at disseminating but also encouraging the use of clinical guidelines should be used for guideline implementation.
Physicians who reported being influenced by or using national guidelines were more likely to be pulmonary or infectious diseases specialists, to spend less time in patient care, to have more CAP inpatients per year, to spend more time reading journals and participating in CME activities, to rate themselves higher on the intellect scale, and to practice at large university/teaching hospitals. Physicians with accurate knowledge about whether a local guideline existed were more likely to practice at large university/teaching hospitals and to see more CAP inpatients per year. However, in contrast to our findings for national guidelines, use of local guidelines was associated with spending more time in patient care. Finally, generalists and physicians who had more positive attitudes about guidelines in general were more likely to use local guidelines.
A central theme of the multivariate findings is that greater reported use and influence of ATS guidelines is independently associated with practice as a pulmonary/infectious diseases specialist, and fewer clinical responsibilities. This is not necessarily surprising given that these specialists were responsible for developing national CAP guidelines, and may be more likely to encounter information about national guidelines in specialty journals or CME. In contrast, it was generalists with more clinical responsibilities who were more likely to use locally developed guidelines. It is possible that physicians who are nonspecialists and more likely to be the attending physician for patients with CAP are more likely to be the target of local—as opposed to national—guideline dissemination efforts. Other studies of guideline implementation suggest that physicians who have less expertise or experience caring for a certain patient population are more likely to rely on guidelines.28,29
These findings could provide an alternative explanation for higher reported local guideline use among generalists as compared to medical specialists. It is important to note that in certain circumstances, noncompliance may be the most appropriate course of action. Since we relied on self-reports of behavior, we have no way to know the extent to which the patients cared for by pulmonary/infectious diseases physicians were similar to those treated by generalists or other specialists or whether they were equally appropriate for guideline-recommended care.
The strengths of our study are that our sample included a broad range of medical specialties and a diverse group of hospitals, which suggests that these results may be generalizable. In addition, the dual focus on national and local guidelines provides a level of specificity about guideline awareness and use not available from previous studies. However, the cross-sectional, self-report nature of the data limit our ability to draw firm conclusions about predictive links between physician and hospital characteristics, and actual physician behavior related to guideline compliance. It is conceivable that physicians manage patients in a manner consistent with guidelines without specific knowledge of the guideline principles. Another potential limitation is that because our survey was conducted prior to the full dissemination of other guidelines (e.g., IDSA), we focused solely on ATS guidelines. However, the current availability of multiple published guidelines could potentially lead to even greater lack of awareness of guidelines than was found in this study, and/or confusion over which guideline to use.
In conclusion, the low levels of reported guideline use —even among physicians who are familiar with the guidelines—and differences in the patterns of awareness and reported use for pulmonary/infectious diseases specialists versus generalists indicate that intensive and targeted intervention efforts may be necessary to encourage compliance with clinical guidelines for the management of CAP.
Table 3.
Factors Related to Physician Knowledge About Whether a Local Community-acquired Pneumonia Guideline Existed
| Hospital and Physician Characteristics* | Total N | Not Accurate† | Accurate† | Bivariate P Value | Multiple Stepwise Odds Ratio (95% CI)‡ |
|---|---|---|---|---|---|
| Hospital characteristics | |||||
| Hospital size, % | .32 | ||||
| Small (<200 beds) | 2 | 0 | 100 | ||
| Medium (200–500 beds) | 236 | 56 | 44 | ||
| Large (>500 beds) | 83 | 63 | 37 | ||
| Teaching status, % | .04 | – | |||
| Teaching | 218 | 61 | 39 | ||
| Nonteaching | 103 | 50 | 50 | ||
| Physician characteristics | |||||
| Specialty, % | .35 | ||||
| Generalist | 251 | 57 | 43 | ||
| Pulmonary or ID specialist | 28 | 50 | 50 | ||
| Other specialist | 42 | 67 | 33 | ||
| Direct inpatient care, h/wk ± SD | 314 | 14.4 ± 14.3 | 16.5 ± 14.5 | .04 | – |
| Direct outpatient care, h/wk ± SD | 315 | 28.5 ± 14.5 | 29.3 ± 15.5 | .83 | |
| Total nonpatient activities, h/wk ± SD | 287 | 11.5 ± 11.7 | 12.8 ± 14.6 | .66 | |
| Percent of total time on inpatient care ± SD | 303 | 25.0 ± 22.6 | 27.8 ± 21.7 | .06 | – |
| Total work, h/wk ± SD | 305 | 55.2 ± 12.7 | 57.9 ± 13.2 | .06 | 1.01* (1.00 to 1.02) |
| Number of CAP inpatients per year ± SD | 315 | 18.4 ± 25.4 | 25.3 ± 32.3 | .02 | – |
| Journal reading, % | .82 | ||||
| Do not read or skim articles | 158 | 58 | 42 | ||
| Read ≥1 article in most issues | 158 | 57 | 43 | ||
| Quality assurance hours, % | .22 | ||||
| <1 h per month | 188 | 54 | 46 | ||
| ≥1 h per month | 131 | 62 | 38 | ||
| Intellect personality score (0–4 = high) ± SD | 288 | 2.81 ± 0.56 | 2.86 ± 0.50 | .51 | |
| Attitudes about guidelines (0–40 = pos) ± SD | 312 | 17.0 ± 5.09 | 17.9 ± 4.63 | .09 | 1.06** (1.01 to 1.11) |
| Risk avoidance (0–4 = high) ± SD | 314 | 3.62 ± 0.85 | 3.81 ± 0.80 | .06 | 1.29* (0.96 to 1.72) |
Gender, age, years since graduation from medical school, university hospital affiliation, percent of income from salary, total patient care (h/wk), continuing medical education hours per month, and personality trait “agreeable” were not significantly related to awareness and use of guidelines or any of the other guideline outcomes presented in Tables 1 and 2–4.
Percentages are calculated within each row.
P value < .10 but ≥ .05,
P value < .05 but ≥ .01,
P value < .01.
CAP, community-acquired pneumonia.
Acknowledgments
Supported in part by AHRQ Grant #RO1HS08282.
REFERENCES
- 1.Wennberg JE, Freeman JL, Culp WJ. Are hospital services rationed in New Haven or over-utilised in Boston? Lancet. 1987;1:1185–9. doi: 10.1016/s0140-6736(87)92152-0. [DOI] [PubMed] [Google Scholar]
- 2.Meehan TP, Fine MJ, Krumholz HM, et al. Quality of care, process and outcomes in elderly patients with pneumonia. JAMA. 1997;278:2080–4. [PubMed] [Google Scholar]
- 3.McCormick D, Fine MJ, Coley CM, et al. Variation in length of hospital stay in patients with community-acquired pneumonia: are shorter stays associated with worse medical outcomes? Am J Med. 1999;107:5–12. doi: 10.1016/s0002-9343(99)00158-8. [DOI] [PubMed] [Google Scholar]
- 4.Fine MJ, Singer DE, Phelps Al, et al. Differences in length of hospital stay in patients with community-acquired pneumonia: a prospective four-hospital study. Med Care. 1993;31:371–80. doi: 10.1097/00005650-199304000-00008. [DOI] [PubMed] [Google Scholar]
- 5.Wickizer TM. The effect of utilization review on hospital use and expenditures: a review of the literature and an update on recent findings. Med Care Rev. 1990;47:327–63. doi: 10.1177/107755879004700303. [DOI] [PubMed] [Google Scholar]
- 6.Halm EA, Fine MJ, Marrie TJ, et al. Time to clinical stability in patients hospitalized with community-acquired pneumonia: implications for practice guidelines. JAMA. 1998;279:1452–7. doi: 10.1001/jama.279.18.1452. [DOI] [PubMed] [Google Scholar]
- 7.Niederman MS, Bass JB, Jr, Campbell GD, et al. Guidelines for the initial management of adults with community-acquired pneumonia: diagnosis, assessment of severity, and initial antimicrobial therapy. American Thoracic Society. Medical Section of the American Lung Association. Am Rev Respir Dis. 1993;128:1418–26. doi: 10.1164/ajrccm/148.5.1418. [DOI] [PubMed] [Google Scholar]
- 8.Niederman MS, Mandell LA, Anzueto A, et al. Guidelines for the management of adults with community-acquired pneumonia. Diagnosis, assessment of severity, antimicrobial therapy, and prevention. Am J Respir Crit Care Med. 2001;163:1730–54. doi: 10.1164/ajrccm.163.7.at1010. [DOI] [PubMed] [Google Scholar]
- 9.Bartlett JG, Breiman RF, Mandell LA, et al. Community-acquired pneumonia in adults: guidelines for management. The Infectious Diseases Society of America. Clin Infect Dis. 1998;26:811–38. doi: 10.1086/513953. [DOI] [PubMed] [Google Scholar]
- 10.Bartlett JG, Dowell SF, Mandell LA, et al. Practice guidelines for the management of community-acquired pneumonia in adults. Infectious Diseases Society of America. Clin Infect Dis. 2000;31:347–82. doi: 10.1086/313954. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Davis DA, Thomson MA, Oxman AD, Haynes RB. Changing physician performance: a systematic review of the effect of continuing medical education strategies. JAMA. 1995;274:700–5. doi: 10.1001/jama.274.9.700. [DOI] [PubMed] [Google Scholar]
- 12.Grimshaw JM, Russell IT. Effect of clinical guidelines on medical practice: a systematic review of rigorous evaluations. Lancet. 1993;342:317–22. doi: 10.1016/0140-6736(93)92244-n. [DOI] [PubMed] [Google Scholar]
- 13.Wells KB, Sherbourne C, Schoenbaum M, et al. Impact of disseminating quality improvement programs for depression in managed primary care: a randomized controlled trial. JAMA. 2000;283:212–20. doi: 10.1001/jama.283.2.212. [DOI] [PubMed] [Google Scholar]
- 14.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. Can Med Assoc J. 1997;157:408–16. [PMC free article] [PubMed] [Google Scholar]
- 15.Pathman DE, Konrad TR, Freed GL, Freeman VA, Koch GG. The awareness to adherence model of steps to clinical guideline compliance. Med Care. 1996;34:873–89. doi: 10.1097/00005650-199609000-00002. [DOI] [PubMed] [Google Scholar]
- 16.Agency for Health Care Policy and Research National Inpatient Statistics. 1996 99–0034:1999. [Google Scholar]
- 17.Fine MJ, Stone RA, Singer DE, et al. Processes and outcomes of care for patients with community-acquired pneumonia: results from the Pneumonia Patient Outcomes Research Team (PORT) Cohort Study. Arch Intern Med. 1999;159:970–80. doi: 10.1001/archinte.159.9.970. [DOI] [PubMed] [Google Scholar]
- 18.Gilbert K, Gleason P, Singer DE, et al. Variation in antimicrobial use and cost in more than 2,000 patients with community-acquired pneumonia. Am J Med. 1998;104:17–27. doi: 10.1016/s0002-9343(97)00274-x. [DOI] [PubMed] [Google Scholar]
- 19.Gleason PP, Kapoor WN, Stone RA, et al. Medical outcomes and antimicrobial costs with the use of the American Thoracic Guidelines for outpatients with community-acquired pneumonia. JAMA. 1997;278:32–9. [PubMed] [Google Scholar]
- 20.Weingarten S. Translating practice guidelines into patient care: guidelines at the bedside. Chest. 2000;118:4–7. doi: 10.1378/chest.118.2_suppl.4s. [DOI] [PubMed] [Google Scholar]
- 21.Halm EA, Switzer GE, Mittman BS, et al. What factors influence physicians’ decisions to switch from intravenous to oral antibiotics for community-acquired pneumonia? J General Intern Med. 2001;16:599–605. doi: 10.1046/j.1525-1497.2001.016009599.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Fine MJ, Stone RA, Lave JR, et al. Implementation of an evidence-based medical practice guideline to reduce duration of intravenous antibiotic therapy and length of stay for patients hospitalized with community-acquired pneumonia: results of a randomized controlled trial. Am J Med. (accepted, under revision). [DOI] [PubMed]
- 23.Jackson DN. Jackson Personality Inventory Manual. Goshen, NY: Research Psychologists Press; 1975. [Google Scholar]
- 24.Goldberg LR. The development of markers for the big five factor structure. Psychol Assessment. 1992;4:26–42. [Google Scholar]
- 25.Tunis SR, Hayward RS, Wilson MC, et al. Internists’ attitudes about clinical practice guidelines. Ann Intern Med. 1994;120:956–63. doi: 10.7326/0003-4819-120-11-199406010-00008. [DOI] [PubMed] [Google Scholar]
- 26.Tunis SR, Hayward RS, Wilson MC, et al. Internists’ attitudes about clinical practice guidelines. Ann Intern Med. 1994;120:956–63. doi: 10.7326/0003-4819-120-11-199406010-00008. [DOI] [PubMed] [Google Scholar]
- 27.Cabana MD, Rand CS, Powe NR, et al. Why don’t physicians follow clinical practice guidelines? A framework for improvement. JAMA. 1999;282:1458–65. doi: 10.1001/jama.282.15.1458. Review. [DOI] [PubMed] [Google Scholar]
- 28.Halm EA, Atlas SJ, Borowsky LH, et al. Change in physician knowledge and attitudes after implementation of a pneumonia practice guideline. J Gen Intern Med. 1999;14:688–94. doi: 10.1046/j.1525-1497.1999.03469.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Mittman BS, Tonesk X, Jacobson PD. Implementing clinical practice guidelines: social influences strategies and practitioner behavior change. Qual Rev Bull. 1992;18:413–22. doi: 10.1016/s0097-5990(16)30567-x. [DOI] [PubMed] [Google Scholar]