Access the full text.
Sign up today, get DeepDyve free for 14 days.
ObjectiveTo compare outcomes for patients with acute low back pain who received care from practitioners with different self-confidence scores on a 4-item scale.DesignCross-sectional survey of practitioners. Prospective cohort study of patient outcomes.SettingPrivate practices and a group model health maintenance organization.ParticipantsOne hundred eighty-nine practitioners, including private practice traditionally trained medical physicians, chiropractors, and physicians in a group model health maintenance organization, who were randomly chosen from practices across the state of North Carolina. These practitioners enrolled 1633 patients with acute low back pain into a prospective cohort study.MethodsThe practitioner survey contained 10 questionnaire items that measured aspects of practitioner confidence and attitudes in assessing and treating patients with low back pain. Patients were interviewed by telephone after the initial office visit and at 2, 4, 8, 12, and 24 weeks, or until complete recovery, whichever came first.ResultsOf 189 study practitioners, 95% responded to the survey. A 4-item scale, shown by factor analysis to describe practitioners' self-confidence, demonstrated good internal consistency among physicians and chiropractors. Chiropractors had significantly stronger self-confidence scores than physicians. Among patients of primary care physicians and chiropractors, those who received care from practitioners with stronger self-confidence scores did not differ in the time to functional improvement, overall patient satisfaction, or their perception of the completeness of care.ConclusionThe level of practitioner self-confidence, as measured by a 4-item scale, did not predict patient outcomes in the treatment of acute low back pain.BACK PAIN is one of the most common clinical problems encountered by primary care practitioners. The lifetime prevalence of at least 1 episode of low back pain (LBP) has been estimated to be 80%.Growing concern over the costs and consequences of back pain disability has led to research aimed at a better understanding of the assessment and treatment of this common problem.A biopsychosocial approach to understanding LBP syndromes suggests that attention to psychological and social aspects of the illness as well as the physician-patient relationship may be important components of successful treatment.In this context, the quality and content of physician-patient interactions, or clinical encounters, may influence the patient outcome. Practitioners who are comfortable and confident in their clinical abilities with LBP and who recognize the possible therapeutic value of nontechnological aspects of the clinical encounter may be more effective in treating patients with this condition.Chiropractors approach LBP diagnosis and treatment from a different theoretical perspective than do physicians.Chiropractors receive 4 years of training focused on the treatment of musculoskeletal problems. In clinical practice, they use various manual techniques that are often referred to as "spinal manipulation." Although back pain is one of the most common problems seen by primary care physicians, they may see only a few cases of it in a week. Most physicians rely on medication and exercises rather than spinal manipulation as the first line of treatment of LBP.Cherkin et alhave described differences between family physicians and chiropractors in their level of confidence and comfort in managing patients with back pain. In general, chiropractors reported higher levels of comfort and confidence in their ability to positively affect the outcome of patients with LBP. In their study, Cherkin et al did not assess patient outcomes, but they suggested that the possible effects of practitioner confidence on the cost and outcome of treatment for LBP should be evaluated.In a study of 18 physicians and 3 physician assistants in a health maintenance organization primary care clinic, Bush et aldescribed a practitioner confidence scale derived from questionnaire items measuring confidence and attitudes when diagnosing and treating LBP. A relationship was reported between scores on a 6-item confidence scale for each provider and only 1 type of patient satisfaction (satisfaction with information received from the provider). Conclusions from the data were limited by the small number of practitioners included in the study, all of whom practiced in the same primary care clinic.We applied the measures reported by Bush et alto a larger, diverse sample of physicians and chiropractors who participated in the North Carolina Back Pain Project to assess aspects of practitioners' confidence in diagnosing and treating patients with acute LBP. We report the effect of practitioners' self-confidence levels on the outcomes of patients in a prospective cohort study who sought care from primary care physicians and chiropractors for acute LBP.PARTICIPANTS AND METHODSThe context of this study was the North Carolina Back Pain Project, a large prospective cohort study examining the process and outcomes of care for acute LBP in community settings across the state. Detailed descriptions of the methods used for sampling and outcome measurement have been published elsewhere.PRACTITIONERSPractitioners were randomly selected from across the state of North Carolina using medical and chiropractic licensure files from 6 strata: urban and rural primary care physicians, urban and rural chiropractors, orthopedic surgeons, and physicians in a group model health maintenance organization. Practitioners were eligible for the study if they were in the designated specialties, practiced in ambulatory care more than half the time, and saw patients with acute LBP as part of their practice without previous referral. Participants completed a mailed questionnaire that included 10 Likert scale items developed by Bush et alregarding self-confidence and attitudes toward dealing with patients with LBP (Table 1). Because this was an observational cohort study, practitioners were given no educational materials and were instructed not to change their usual patterns of care. All data collection beyond the index visit for back pain was performed by study personnel over the telephone, thus limiting provider contact with the study personnel.Table 1. 10 Likert Scale Items Assessing Practitioners' Attitudes Toward Treating Low Back Pain (LBP)*†See table graphicResponses of primary care physicians to the 10 questionnaire items were assessed using factor analysis techniques. Factor analysis examines the covariance between items and helps to determine the groupings of items that correspond with 1 or more underlying constructs or latent variables. Graphs of eigenvalues (scree plots) were generated to determine the number of factors to retain in factor analysis.Interitem correlations and appropriate groupings of items were determined using principal components analysis with varimax rotation. An item was included in a grouping if it correlated most strongly with the factor (factor loading >0.40) and did not have a factor loading greater than 0.30 with any other factor. This factor analytic process was then used to evaluate questionnaire responses from participating chiropractors. This process resulted in a grouping of 4 items that appeared to measure underlying aspects of practitioners' self-confidence.Reliability or internal consistency of the resulting scale was assessed with the Cronbach coefficient α. This is defined as the proportion of a scale's total variance that is attributable to a common underlying source, presumably the latent variable of interest (practitioners' self-confidence).A score from the 4-item scale was calculated for each practitioner by adding the Likert score from each of the 4 questionnaire items (ranges, 1-5 for each questionnaire item and 4-20 for the 4-item scale). Distributions of self-confidence scale scores for physicians and chiropractors were compared using the Mann-Whitney Utest. This nonparametric test was chosen because of the nonnormal distribution of the scores and the ordinal nature of the data.Factor analysis, the Cronbach α, and statistical comparisons for practitioner self-confidence measures were all calculated using commercially available statistical software.PATIENTSConsecutive patients with acute LBP were invited into the study by practitioners. Patients were eligible for the study if they had LBP for less than 10 weeks, had not previously sought care for the episode, never had a back operation, did not have a history of metastatic malignancy, were not pregnant, owned a telephone, and spoke English. Patients were interviewed by telephone after the initial office visit and at 2, 4, 8, 12, and 24 weeks or until they had full recovery from the episode of acute LBP.The primary patient outcome was the date of return to functioning at the same level as before the episode of acute LBP. We also asked at each interview whether patients were "completely better." Satisfaction with care received was measured at the interview when the patient reported complete recovery or at the 24-week interview, whichever came first. Patient satisfaction was measured using questionnaire items adapted from a previous study of LBP.We examined outcomes of patients who sought care from primary care physicians and of patients of chiropractors. (A detailed analysis of the differences in outcomes between the 6 different strata of practitioners has been published elsewhere.) Primary care physicians and chiropractors were divided into groups that most closely approximated tertiles of high, moderate, and lower practitioner self-confidence. We examined bivariate relationships between the levels of practitioner self-confidence scores and patient outcomes and demographic characteristics. We used χ2analysis to compare categorical variables. Kaplan-Meier plots and log-rank tests were used to assess the time to functional recovery. A Cox proportional hazards regression model was used to estimate the effect of practitioner confidence on the time to functional recovery, with adjustments for potential confounding variables. Hazard ratios with 95% confidence limits were calculated from proportional hazards models. Statistical software was used for all outcome data analysis.RESULTSOf 189 practitioners who participated and recruited subjects into the North Carolina Back Pain Project, 179 returned practitioner questionnaires (overall return rate, 94.7%), and 162 (85.7%) of these completed all 10 self-confidence and attitude questionnaire items (physicians, 107; chiropractors, 55). The results of principal components factor analysis with varimax rotation are shown in Table 2. The strongest grouping of 4 items had primary factor loadings greater than 0.68 for each item (items A, B, C, and J). A second grouping showed smaller loadings (0.49-0.72) on factor 2 (items H, F, D, and G).Table 2. Loadings of Each Questionnaire Item After Factor Analysis With Varimax Rotation*See table graphicThe content of the first 4-item scale (items A, B, C, and J) described the underlying construct of primary interest: practitioner self-confidence in the diagnosis and management of LBP. The content of the second grouping of 4 items (H, F, D, and G) described practitioner attitudes toward patients. The remaining 2 items (I and E) focused on specific outcomes of the natural history and treatment of acute LBP (progression of acute to chronic pain and patient satisfaction with treatment). Factor analysis of the chiropractors' responses resulted in similar groupings of items (self-confidence and attitudes). Unlike for the physicians, however, the chiropractors' responses on item I ("Patients are satisfied with my care") also had a strong correlation with the 4 self-confidence items.We assessed the self-confidence scores for physicians and chiropractors using the 4-item scale that was common to both groups in the factor analysis. This scale had a Cronbach coefficient α of .73 for physicians and .70 for chiropractors. Responses from the 4 items were summed for each practitioner to give a self-confidence score. The distribution of practitioner self-confidence scores for chiropractors showed significantly greater self-confidence when compared with that of all physicians (Figure 1; Mann-Whitney Utest, z=−7.1, P=.001).Figure 1.Percentages of physicians and chiropractors at each level of the self-confidence score. The score for each practitioner is a sum of the Likert score of the 4 self-confidence items (A, B, C, and J) after reverse-scoring item A (Table 2). Note that lower scores indicate greater self-confidence in diagnosing and managing low back pain.TIME TO FUNCTIONAL RECOVERYPatients seen by primary care physicians with stronger self-confidence scores did not recover faster than patients of physicians with weaker scores. A Kaplan-Meier analysis of patients seen by primary care physicians (Figure 2) showed no significant difference in the number of days to functional recovery (a return to a functional status similar to that before the onset of LBP). Additional Kaplan-Meier plots using different cutoff points for the self-confidence scale also failed to show differences in the time to functional recovery for patients of highly confident primary care physicians. A Cox proportional hazards model adjusting for patient sex, age, race, general health, and baseline level of disability from LBP (the Roland-Morris scale) also showed no significant difference in time to functional recovery for patients seen by primary care physicians with strong (highest tertile) self-confidence scores (hazards ratio, 0.99; 95% confidence limits 0.90, 1.11).Figure 2.Kaplan-Meier plot of the number of days to functional recovery for patients with acute low back pain seen by primary care physicians. Dotted line indicates patients (n=203) seen by physicians with stronger self-confidence scores (≤7, Table 2); solid line, patients (n=413) seen by physicians with weaker self-confidence scores (P>.70).A similar analysis of the time to functional recovery for patients of chiropractors also showed no significant effect of practitioner self-confidence. A Kaplan-Meier analysis (Figure 3) showed no difference in the time to functional recovery for patients seen by chiropractors with the strongest self-confidence score. After adjusting for the patient characteristics noted above in a Cox proportional hazards model, seeing a chiropractor with the strongest self-confidence score was not associated with a faster recovery (hazards ratio, 1.04; 95% confidence limits, 0.95, 1.14).Figure 3.Kaplan-Meier plot of the number of days to functional recovery for patients with acute low back pain seen by chiropractors. Dotted line indicates patients (n=207) seen by chiropractors with the strongest self-confidence score (4 on the 4-item scale, Table 2); solid line, patients (n=344) seen by chiropractors with weaker self-confidence scores (P>.60).PRACTITIONER SELF-CONFIDENCE AND PATIENT SATISFACTIONWe have previously reported significantly higher levels of patient satisfaction and perception of completeness of care for patients of chiropractors in these patient cohorts.Within groups of patients seen by physicians and chiropractors, there were no significant differences in measures of patient satisfaction across 3 levels of practitioner self-confidence (Table 3). There was a nonsignificant trend for patients of primary care physicians with weaker self-confidence to be more satisfied with the information they were given by their physician and to be more satisfied with overall treatment results.Table 3. Patient Satisfaction With Their Low Back Pain (LBP) Treatment*See table graphicThere are also previously reported differences between patients of physicians and chiropractors in their perception of the completeness of care from these patient cohorts.Within groups of practitioners, however, there were no significant differences across self-confidence levels in patients' perception of the completeness of care (Table 4).Table 4. Patients' Perception of Care for Low Back Pain*See table graphicCOMMENTStriking differences between self-confidence scores of physicians and chiropractors in our sample are consistent with previous findings. Cherkin et alfound that 42% of physicians felt poorly prepared to manage LBP when they first entered practice, compared with only 15% of chiropractors. That chiropractors are more confident of their skills and have patients with higher levels of satisfaction has led to hypotheses that practitioner confidence may affect a variety of outcomes among patients with LBP.In data previously published from the North Carolina Back Pain Project,patients had similar outcomes in time to the improvement of function and the resolution of pain no matter which type of practitioner they saw. The results of our analysis of practitioner self-confidence suggests that receiving care from a more confident practitioner does not lead to a faster return to baseline functioning. This conclusion is strengthened by the consistent findings across confidence groups in the separate analyses of patients of primary care physicians and those of chiropractors. In addition, our results did not support the hypothesis that greater practitioner confidence leads to more satisfied patients. Does the level of practitioner confidence really not matter to patients with back pain? Possible limitations and strengths of our study must first be considered.The scope of our measure of practitioner confidence was limited but showed evidence of good reliability in our sample. The Cronbach α is a measure of internal consistency or reliability among the items of the scale. An α close to 1.0 is desirable, but an α in the range of .70 to .80 is evidence of acceptable reliability.The α scores for the self-confidence scale in our study were within an acceptable range of reliability for research scales and are similar to those reported for other practitioner self-confidence scales in diagnosing and treating depression (5-item scale, α=.73)and in counseling patients to make lifestyle changes (18-item scale, α=.77).Evidence of the reliability of the self-confidence scale is strengthened by the similar α scores we found among groups of physicians and chiropractors.The 6-item practitioner confidence scale reported by Bush et alcontained the 4 items in our self-confidence score and 2 other items (D and E, Table 1). These differences in the factor analysis results were most likely due to a larger, more diverse population in our study. The confidence scale of Bush et al was derived from responses of only 21 practitioners, all of whom practiced in the same outpatient setting. Their responses to questionnaire items may have been more homogeneous because of this, and they were more likely to have similar approaches to patient care than the diverse cross section of physicians and chiropractors in our study. Our study population met the suggested minimum for accurate factor analysis of 5 to 10 respondents per item in the questionnaire (at least 50-100 respondents for the 10 items included).The wide dispersion of scores for physicians in our sample is consistent with good discrimination of various levels of self-confidence.Potentially significant aspects of practitioner-patient interactions are not measured by our brief scale, including practitioner communication skills, the amount of time spent with each patient, the amount of hands-on evaluation and treatment, and the level of empathy and emotional support given by the practitioner. Measures of general confidence in clinical skills may not be correlated with practitioners' ability to communicate effectively with patients. These and other psychosocial factors in the physician-patient encounter may affect outcomes, particularly for patients who have chronic pain syndromes.Significant differences in their approach to LBP and practice profiles likely explain higher confidence scores among chiropractors in our sample.Chiropractic is a system of care that emphasizes a hands-on approach to back pain, with detailed clinical examinations and the frequent use of manual therapy techniques. Few medically trained physicians perform spinal manipulation, and they tend to examine patients with acute LBP with a brief screening examination to rule out neurological impairment and underlying systemic disease. (Note: Spinal manipulation is undertaken by doctors of osteopathy, but few practice in North Carolina, and none were included in our study.) Given the differences between chiropractors and physicians in their theoretical perspective, their approach to treatment, and the larger volume of cases of LBP seen by chiropractors, each group is likely to have different underlying paradigms of self-confidence.Differences between patients of physicians and chiropractors in their level of satisfaction with LBP care have been previously reported.Bush et alposited that some of these differences in satisfaction level could be explained by practitioners' confidence and attitudes toward patients with LBP. Our results indicate that practitioners' self-confidence does not explain differences in patient satisfaction between physicians and chiropractors. This conclusion is strengthened by the lack of variation in patient satisfaction across levels of confidence within each practitioner group. Previous multivariate analysis of overall patient satisfaction from the North Carolina Back Pain study has shown that a greater number of office visits and more frequent use of radiographs also do not explain greater satisfaction among chiropractic patients.Other aspects of the process of care and practitioner self-confidence that were not specifically measured in our study may help explain differences between physicians and chiropractors that affect satisfaction and perception of care among their patients with acute LBP.For example, closer attention to history taking, more of a hands-on approach to examination and treatment, giving patients specific diagnostic and explanatory labels for the underlying reason for their pain, and formulating clear plans for active rehabilitation may be keys to raising patients' level of satisfaction with care.CONCLUSIONOur findings are consistent with other published evidence that practitioners have little to offer patients with acute LBP that will substantially alter the course of their illness. The physician's job is to screen for underlying systemic disease as a cause of pain, rule out significant nerve root impingement, assist the patient with pain control, give empathic attention to psychosocial matters associated with the illness, and provide advice for the timely return to usual daily activities. We can successfully complete these tasks with a simple, structured clinical approach to the management of acute LBP.Bolstering our confidence in more detailed diagnostic or therapeutic maneuvers for acute LBP is unlikely to benefit most of the patients we see in primary care practice.BKCypressCharacteristics of physician visits for back symptoms: a national perspective.Am J Public Health.1983;73:389-395.JWFrymoyerWLCats-BarilAn overview of the incidence and costs of low back pain.Orthop Clin North Am.1991;22:263-271.RADeyoJRainvilleDLKentWhat can the history and physical examination tell us about low back pain?JAMA.1992;268:760-765.WEFordyceJABrockwayJBergmanDSpenglerAcute back pain: a control-group comparison of behavioral vs traditional management methods.J Behav Med.1986;9:127-140.TSCareyJGarrettAJackmanCMcLaughlinJFryerDRSmuckerThe outcomes and costs of care for acute low back pain among patients seen by primary care practitioners, chiropractors, and orthopedic surgeons.N Engl J Med.1995;333:913-917.GWaddellA new clinical model for the treatment of low-back pain.Spine.1987;12:632-644.DCCherkinFAMacCornackAOBergManaging low back pain: a comparison of the beliefs and behaviors of family physicians and chiropractors.West J Med.1988;149:475-480.MJBassCBuckLTurnerGDickieGPrattHCRobinsonThe physician's actions and the outcome of illness in family practice.J Fam Pract.1986;23:43-47.DCCherkinFAMacCornackPatient evaluations of low back pain care from family physicians and chiropractors.West J Med.1989;150:351-355.TBushDCherkinWBarlowThe impact of physician attitudes on patient satisfaction with care for low back pain.Arch Fam Med.1993;2:301-305.DHNovackTherapeutic aspects of the clinical encounter.J Gen Intern Med.1987;2:346-355.HVernonChiropractic: a model of incorporating the illness behavior model in the management of low back pain patients.J Manipulative Physiol Ther.1991;14:379-389.RFDevellisScale Development: Theory and Applications.Newbury Park, Calif: Sage Publications; 1991.Not AvailableSAS/STAT User's Guide, Version 6.03.Cary, NC: SAS Institute Inc; 1988.MRolandRMorrisA study of the natural history of back pain, part 1: development of a reliable and sensitive measure of disability in low-back pain.Spine.1983;8:141-144.DSMainULutzJEBarrettJMatthewRSMillerThe role of primary care clinician attitudes, beliefs, and training in the diagnosis and treatment of depression.Arch Fam Med.1993;2:1061-1066.SCThompsonLSchwankovskyJPittsCounseling patients to make lifestyle changes: the role of physician self-efficacy, training and beliefs about causes.Fam Pract.1993;10:70-75.JLCoulehanChiropractic and the clinical art.Soc Sci Med.1985;21:383-390.PCurtisGBoveFamily physicians, chiropractors, and back pain.J Fam Pract.1992;35:551-555.RADeyoAKDiehlPatient satisfaction with medical care for low-back pain.Spine.1986;11:28-30.SBigosOBowyerGBrainClinical Practice Guideline Number 14: Acute Low Back Pain in Adults.Rockville, Md: US Dept of Health and Human Services, Agency for Health Care Policy and Research; 1994. AHCPR publication 95-0643.Accepted for publication April 28, 1997.This study was supported in part by grant HS06664 from the Agency for Health Care Policy and Research, Rockville, Md, National Health Services Award grant 5-T32-PE-14001 from the US Department of Health and Human Services, Washington, DC, and a primary care research fellowship from the Department of Family Medicine, University of North Carolina at Chapel Hill (Dr Smucker).Corresponding author: Douglas R. Smucker, MD, Department of Family Medicine, University of Cincinnati College of Medicine, PO Box 670582, Cincinnati, OH 45267-0582.
Archives of Family Medicine – American Medical Association
Published: Jun 1, 1998
Access the full text.
Sign up today, get DeepDyve free for 14 days.