Triage-based care of people with Back Pain: STarT Back or Start diagnosing? An observational study.

ObjectivesBack pain is a massive public health problem. The STarT Back Screening Tool (SBST) was developed for use in primary care to triage people with lumbar pain, classifying them as low, medium or high "risk" of prolonged symptoms. This classification guides non-surgical interventions including manual treatments, exercise and cognitive behavioural therapy. Claims suggest SBST brings generic health and cost benefits. National guidance recommends STarT Back is used at the first primary care consultation but can be used at any stage. For SBST to be an effective triage tool it should distinguish structural from non-structural pain. We tested this requirement in consecutive people referred to a single triage practitioner, hypothesising it was not possible conceptually. DesignAn observational study of the relationship between routine, prospectively collected triage data and diagnosis. SettingA secondary care spinal triage service based in a teaching hospital. ParticipantsWe studied consecutive referrals with lumbar pain triaged by a single extended scope practitioner (ESP) over 22 months (Nov 2015-Sept 2017). Main Outcome MeasuresSBST and pain visual analogue scores (VAS: 0-10) were collected at the initial consultation. We compared data for people with and without surgically remedial lesions. Results1041 people were seen (61% female, mean age 53), n=234 (28%) had surgically amenable explanations for pain. People with surgical lesions were older (58 v 51yrs), more likely male (48 v 35%) and had higher VAS scores (6.8 v 6.1). Surgery and non-surgery subgroups had similar SBST total and domain score distribution profiles. The surgery subgroup had less low risk (9%v21%) and more high risk (37% v 30%) classified people. ConclusionSBST scores did not differentiate surgical from non-surgical pathologies. It seems unlikely that symptom questionnaires can estimate prognosis accurately unless everyone has the same diagnosis, not just the same symptom. Diagnosis, rather than questionnaire scores, should guide treatment and inform prognosis. O_TEXTBOXSummary Box What is already known on this topic?The symptom of low back pain is a common cause of disability worldwide. The majority if people with low back pain probably do not have a structural problem in their lumbar spine to explain the extent of their disability. National guidelines throughout the world attempt to facilitate the identification and treatment of people with, "non-specific low back pain", and prescribe treatment for people given this label. What this study adds?It seems unlikely that symptom questionnaires can estimate prognosis accurately unless everyone has the same diagnosis, not just the same symptom. The practice of recommending treatment and prognosticating in the absence of a diagnosis needs further scrutiny. C_TEXTBOX