Subacute and Chronic, Non-specific Back and Neck Pain: Cognitive-behavioral Rehabilitation vs. Traditional Primary Care

Overview

BACKGROUND Non-specific back and neck pain (BNP) dominates sick-listing. A program of cognitive-behavioral rehabilitation for subacute and chronic BNP was compared, with 18-month follow-up, with traditional primary care concerning sick-listing and health-care visits. METHODS After stratification to age (44 (years) and younger/45 and older) and subacute/chronic BNP (= full-time sick-listed 43-84/85-730 days respectively), 125 primary-care patients were randomized to a rehabilitation center or continued health-center care. Outcome measures were Return-to-work (=the proportion who regained work ability for at least 30 consecutive days), the proportion with Work ability at different time points, Total sick-listing (expressed in whole days) and the total number of Visits (to physicians, physiotherapists etc.) 1-18 months and corresponding six-month periods. For the analyses were used t-test, z-test, generalized estimating equations and a mixed, linear model.

Numbers within parenthesis refers to the place of order in the citation list and within brackets in the link list below. B A C K G R O U N D In Sweden, as all over the industrial world, unspecific back and neck pain (BNP) dominates sick-listing (1). Primary care is the appropriate source of treatment of most patients with BNP (2). However, the Swedish traditional primary care lacks the capacity of such an assignment \[1\]. While the number of practicing physicians is in line with OECD standards, Sweden has, relatively seen, few physicians within primary care. Our overall aim was to compare a program of cognitive-behavioral-rehabilitation at a rehabilitation center for patients with subacute and chronic BNP with traditional primary care. The specific aim of this study was to answer the question: Will the outcome, with an 18-month follow-up, differ concerning sick-listing and number of health-care visits? M E T H O D S PARTICIPANTS: One-hundred-and-twenty-five patients were recruited by 42 family doctors at 12 health centers. INCLUSION AND EXCLUSION CRITERIA: See below INTERVENTIONS: Cognitive-behavioral rehabilitation: The medical, biomechanical and psychosocial obstacles to working were mapped out. A physiotherapist let the patient into graded activity (3). A behaviorist offered cognitive-behavioral therapy. A health adviser taught applied relaxation (4). A physician prescribed medicine when needed. Then the individual management was replaced by team conferences. A rehabilitation plan was drawn up. The patient gradually returned to work. The end of rehabilitation came when the final aim was achieved or when it was clear that work ability would not be attained. Participation in the rehabilitation group did not exclude the patient from seeking other care also. Traditional primary care: The hub of Swedish primary care is the health centres. Besides family doctors, their staff consists of, among others, physiotherapists and social workers. In total, the health centers of this study engaged 84 family doctors and served a population of 148 000 individuals, i.e. slightly less than 0.6, as compared with an OECD-average of 0.8, family doctors/1000 population \[2\]. Participation in the health-center group excluded the patient from turning to the rehabilitation center but not from any other health-care, for example, orthopedist consultation. DATA COLLECTION: Sick-listing data were provided by the Stockholm County Social Insurance Agency. Data of the treatment at the rehabilitation center were collected from its medical journals. As to the rest, health-care data were obtained from follow-up forms. OUTCOME MEASURES: See below. POWER CALCULATION: It originated from a retrospective preliminary study of 172 patients at the rehabilitation centre and from a forecast of the probability of ever regaining work ability for patients with full-time sick-listing for back pain in traditional care (5). The proportion of patients with any degree of work ability at the end of the rehabilitation was 76% and, for the patients with subacute and chronic BNP, 89% and 73% respectively. The average probability of regaining work ability in the case of continued management within traditional care was calculated for each one of the 172 patients according to their period of sick-listing at the start of the rehabilitation and was on average 49% as to be compared with the 76% who really regained work ability. The smallest difference that we wished to demonstrate was 22%. With a significance level of .05 and a power of 80% 154 patients had to be included, and to allow a certain dropout, 170 patients. PREMATURE STOP OF RECRUITMENT: The recruitment was discontinued in January 2004 at 125 patients. The reason was that in April 2004 a large back-rehabilitation centre started in a neighboring municipality. We presumed that many of the planned future patients of the health-center group would be referred to that centre and get a management that could no longer be defined as traditional primary care. INCLUSION PROCEDURE: A patient who fulfilled the criteria and agreed to participate was interviewed by telephone by a research assistant. The patients who still qualified saw the assistant at the health center and went through a start form. Then the assistant carried out a 10-test package, including a lift test. The reliability of that test procedure was confirmed in a separate study (6). Then the randomization was made with stratifications to age 44 and younger and 45 and older, and to subacute BNP and chronic BNP, i.e. full-time sick listed 43-84 and 85-730 days respectively. The stratifications were made out from the results of the preliminary study. Randomization envelopes were used. The patients of the rehabilitation group started the program at the centre within one week. The patients of the health-center group continued at their health-centers (= traditional primary care). FOLLOW-UP: Six, 12 and 18 months after inclusion, the patients were sent follow-up forms. ANALYSES AND STATISTICS: A p-value\<.05 was considered statistically significant. For interval data were used t-test, ordinal data Wilcoxon rank-sum test and proportions z-test. Total sick-listing as well as Visits for the pre- and post-inclusion six-months periods was compared with a mixed-linear model (7).