Who's in charge? Challenges in evaluating quality of primary care treatment for low back pain

    loading  Checking for direct PDF access through Ovid


Rationale and objectives

Low back pain (LBP) is a common condition with frequent health care visits and work disability. Quality improvement efforts in primary care focused on guidelines adherence, provider selection and education, and feedback on appropriateness of care. Such efforts can only succeed if a health care provider is in charge of care over a substantial period. This study was conducted to provide insights about actual patterns of provider involvement in LBP care and implications for quality evaluation.


Established primary care patients with occupational LBP and health care covered by a workers' compensation insurer were selected. Primary care physician (PCP) involvement was examined relative to overall health care utilization. Four methods of classifying PCP involvement were used to assess the association between PCP involvement and health care and work disability outcomes over a 2-year follow-up period.


Primary care physician was rarely the sole provider during episodes of occupational LBP. PCP was the initial non-emergency room provider in 55% of cases, and was the most prevalent provider during at least one episode of care in 45% of cases. Different methods of classification led to different conclusions about the association between PCP involvement and work disability or number of health care visits. Multiple providers were involved throughout the clinical course of the small number of cases that accounted for most of the health care visits and work disability; in these cases, the role of PCP in care was difficult to determine.


Administrative data alone are adequate for provider comparisons only in relatively simple cases. Provider comparisons based on initial treating provider likely overstate the importance of early care, particularly in more complex cases. For LBP, quality improvement models based on PCP-directed interventions or reinforcing guideline adherence may not impact outcomes. A patient-centred model may be necessary to achieve outcome improvements.

Related Topics

    loading  Loading Related Articles