STarT Back: The Questionnaire That Sorts Risk, Not Diagnosis

What is the STarT Back Screening Tool?

STarT Back — the capitals stand for Subgroups for Targeted Treatment — is a nine-item questionnaire developed in United Kingdom primary care to identify which patients with low back pain are at risk of persistent disabling pain 1Ref 1Hill JC, Dunn KM, Lewis M, et al. (2008).A primary care back pain screening tool: identifying patient subgroups for initial treatment.The STarT Back Screening Tool's construction and scoring: nine items covering referred leg pain, comorbid pain, two disability items, bothersomeness, catastrophizing, fear, anxiety and depression, with the final five forming a psychosocial subscale; a total range of 0-9 and a subscale range of 0-5; higher scores meaning greater risk of persistent disabling low back pain; the tool being prognostic rather than diagnostic; the three-subgroup cutoffs (0-3 low; >=4 with subscale >=4 high; >=4 with subscale <=3 medium); the development sample of 131 and independent external validation sample of 500 UK primary-care patients; the reported reliability and validity and acceptability; the intended matching of care intensity to subgroup; and the explicit absence of any MCID or MDC for this instrument.. The total runs from 0 to 9, and a higher score means greater risk 1Ref 1Hill JC, Dunn KM, Lewis M, et al. (2008).A primary care back pain screening tool: identifying patient subgroups for initial treatment.The STarT Back Screening Tool's construction and scoring: nine items covering referred leg pain, comorbid pain, two disability items, bothersomeness, catastrophizing, fear, anxiety and depression, with the final five forming a psychosocial subscale; a total range of 0-9 and a subscale range of 0-5; higher scores meaning greater risk of persistent disabling low back pain; the tool being prognostic rather than diagnostic; the three-subgroup cutoffs (0-3 low; >=4 with subscale >=4 high; >=4 with subscale <=3 medium); the development sample of 131 and independent external validation sample of 500 UK primary-care patients; the reported reliability and validity and acceptability; the intended matching of care intensity to subgroup; and the explicit absence of any MCID or MDC for this instrument.. Five of the nine items form a separate psychosocial subscale scored from 0 to 5 1Ref 1Hill JC, Dunn KM, Lewis M, et al. (2008).A primary care back pain screening tool: identifying patient subgroups for initial treatment.The STarT Back Screening Tool's construction and scoring: nine items covering referred leg pain, comorbid pain, two disability items, bothersomeness, catastrophizing, fear, anxiety and depression, with the final five forming a psychosocial subscale; a total range of 0-9 and a subscale range of 0-5; higher scores meaning greater risk of persistent disabling low back pain; the tool being prognostic rather than diagnostic; the three-subgroup cutoffs (0-3 low; >=4 with subscale >=4 high; >=4 with subscale <=3 medium); the development sample of 131 and independent external validation sample of 500 UK primary-care patients; the reported reliability and validity and acceptability; the intended matching of care intensity to subgroup; and the explicit absence of any MCID or MDC for this instrument..

Those two numbers together place a person in one of three groups: low, medium or high risk 1Ref 1Hill JC, Dunn KM, Lewis M, et al. (2008).A primary care back pain screening tool: identifying patient subgroups for initial treatment.The STarT Back Screening Tool's construction and scoring: nine items covering referred leg pain, comorbid pain, two disability items, bothersomeness, catastrophizing, fear, anxiety and depression, with the final five forming a psychosocial subscale; a total range of 0-9 and a subscale range of 0-5; higher scores meaning greater risk of persistent disabling low back pain; the tool being prognostic rather than diagnostic; the three-subgroup cutoffs (0-3 low; >=4 with subscale >=4 high; >=4 with subscale <=3 medium); the development sample of 131 and independent external validation sample of 500 UK primary-care patients; the reported reliability and validity and acceptability; the intended matching of care intensity to subgroup; and the explicit absence of any MCID or MDC for this instrument.. That is the whole output. There is no diagnosis in it, no severity grade, and no statement about what is happening structurally in anyone's spine.

STarT Back is a prognostic instrument. It predicts how things are likely to go, not what is wrong. That distinction does more work than it first appears to, because almost every other form handed out in a back-pain appointment is trying to answer a different question.

It was published in 2008 by Hill and colleagues, developed on a sample of 131 patients and then tested in an independent external validation sample of 500 primary-care patients, with reliability and concurrent and discriminant validity described as good, and with the tool reported as acceptable to both patients and clinicians 1Ref 1Hill JC, Dunn KM, Lewis M, et al. (2008).A primary care back pain screening tool: identifying patient subgroups for initial treatment.The STarT Back Screening Tool's construction and scoring: nine items covering referred leg pain, comorbid pain, two disability items, bothersomeness, catastrophizing, fear, anxiety and depression, with the final five forming a psychosocial subscale; a total range of 0-9 and a subscale range of 0-5; higher scores meaning greater risk of persistent disabling low back pain; the tool being prognostic rather than diagnostic; the three-subgroup cutoffs (0-3 low; >=4 with subscale >=4 high; >=4 with subscale <=3 medium); the development sample of 131 and independent external validation sample of 500 UK primary-care patients; the reported reliability and validity and acceptability; the intended matching of care intensity to subgroup; and the explicit absence of any MCID or MDC for this instrument.. It is short by design: a form that takes two minutes gets filled in, and one that takes fifteen does not.

Risk of what, exactly?

The risk being estimated is the risk of persistent disabling low back pain — pain that is still limiting what you can do months from now 1Ref 1Hill JC, Dunn KM, Lewis M, et al. (2008).A primary care back pain screening tool: identifying patient subgroups for initial treatment.The STarT Back Screening Tool's construction and scoring: nine items covering referred leg pain, comorbid pain, two disability items, bothersomeness, catastrophizing, fear, anxiety and depression, with the final five forming a psychosocial subscale; a total range of 0-9 and a subscale range of 0-5; higher scores meaning greater risk of persistent disabling low back pain; the tool being prognostic rather than diagnostic; the three-subgroup cutoffs (0-3 low; >=4 with subscale >=4 high; >=4 with subscale <=3 medium); the development sample of 131 and independent external validation sample of 500 UK primary-care patients; the reported reliability and validity and acceptability; the intended matching of care intensity to subgroup; and the explicit absence of any MCID or MDC for this instrument.. It is not the risk of a fracture, a tumour, an infection, or a compressed nerve. A high-risk STarT Back result is not a warning that something dangerous has been missed, and a low-risk result is not a clearance certificate.

That second point deserves emphasis, because the two kinds of screening get confused constantly. Screening for a serious underlying cause is done with low back red flags, and the evidence on those is its own uncomfortable subject: a systematic review of red flags for fracture and malignancy found that most individual red flags carry high false-positive rates, although some — older age, prolonged corticosteroid use, and significant trauma when the question is fracture — do meaningfully raise the probability 3Ref 3Downie A, Williams CM, Henschke N, et al. (2013).Red flags to screen for malignancy and fracture in patients with low back pain: systematic review.That screening for serious underlying causes of low back pain is done with red flags rather than a risk score, and that most individual red flags carry high false-positive rates while some — older age, prolonged corticosteroid use, and significant trauma when screening for fracture — meaningfully raise post-test probability.. STarT Back contains none of that. It was never built to.

Severity is a third question again, and it has its own instruments. The Oswestry Disability Index measures how much low-back-related disability a person currently has, across ten fixed sections scored as a percentage 4Ref 4Fairbank JCT, Pynsent PB (2000).The Oswestry Disability Index.That the Oswestry Disability Index is a validated ten-section patient-reported measure of current low-back-pain-related disability scored as a percentage — used here as the contrast case for a severity measure rather than a prognostic one.. Someone can score badly on the Oswestry today and land in the low-risk STarT Back group, because the first form is asking how bad things are now and the second is asking how likely they are to stay that way.

Prognostic means predictive of the future course; diagnostic means identifying the current cause. STarT Back is entirely the first and not at all the second.

What the nine items cover

The nine items span two territories. Four ask about the physical picture: pain referred into the leg, pain elsewhere in the body alongside the back pain, and two items about disability 1Ref 1Hill JC, Dunn KM, Lewis M, et al. (2008).A primary care back pain screening tool: identifying patient subgroups for initial treatment.The STarT Back Screening Tool's construction and scoring: nine items covering referred leg pain, comorbid pain, two disability items, bothersomeness, catastrophizing, fear, anxiety and depression, with the final five forming a psychosocial subscale; a total range of 0-9 and a subscale range of 0-5; higher scores meaning greater risk of persistent disabling low back pain; the tool being prognostic rather than diagnostic; the three-subgroup cutoffs (0-3 low; >=4 with subscale >=4 high; >=4 with subscale <=3 medium); the development sample of 131 and independent external validation sample of 500 UK primary-care patients; the reported reliability and validity and acceptability; the intended matching of care intensity to subgroup; and the explicit absence of any MCID or MDC for this instrument.. The other five ask about the psychological picture: how bothersome the pain has been, catastrophic thinking about it, fear of movement, anxiety, and low mood 1Ref 1Hill JC, Dunn KM, Lewis M, et al. (2008).A primary care back pain screening tool: identifying patient subgroups for initial treatment.The STarT Back Screening Tool's construction and scoring: nine items covering referred leg pain, comorbid pain, two disability items, bothersomeness, catastrophizing, fear, anxiety and depression, with the final five forming a psychosocial subscale; a total range of 0-9 and a subscale range of 0-5; higher scores meaning greater risk of persistent disabling low back pain; the tool being prognostic rather than diagnostic; the three-subgroup cutoffs (0-3 low; >=4 with subscale >=4 high; >=4 with subscale <=3 medium); the development sample of 131 and independent external validation sample of 500 UK primary-care patients; the reported reliability and validity and acceptability; the intended matching of care intensity to subgroup; and the explicit absence of any MCID or MDC for this instrument.. Those five make up the psychosocial subscale.

The tool itself is copyrighted by its developers and is not reproduced here; what follows describes the constructs the items measure, not their wording. Clinics that use it obtain the licensed form, and the scoring instructions come with it.

The split is the design. A questionnaire that only counted physical findings would put everyone with leg pain and widespread symptoms into a high-risk box. A questionnaire that only counted distress would miss the person whose pain genuinely is spreading. STarT Back weights both, then treats the psychosocial cluster as the deciding factor at the top end.

One item repays a second look, because its placement surprises people. Bothersomeness sits in the psychosocial cluster rather than the physical one 1Ref 1Hill JC, Dunn KM, Lewis M, et al. (2008).A primary care back pain screening tool: identifying patient subgroups for initial treatment.The STarT Back Screening Tool's construction and scoring: nine items covering referred leg pain, comorbid pain, two disability items, bothersomeness, catastrophizing, fear, anxiety and depression, with the final five forming a psychosocial subscale; a total range of 0-9 and a subscale range of 0-5; higher scores meaning greater risk of persistent disabling low back pain; the tool being prognostic rather than diagnostic; the three-subgroup cutoffs (0-3 low; >=4 with subscale >=4 high; >=4 with subscale <=3 medium); the development sample of 131 and independent external validation sample of 500 UK primary-care patients; the reported reliability and validity and acceptability; the intended matching of care intensity to subgroup; and the explicit absence of any MCID or MDC for this instrument.. That is not a filing error. How bothersome pain is describes what the pain is doing to your life and your attention, which is a different quantity from how intense it is — and it is the first of the five that most people recognise in themselves.

One practical consequence follows from how short it is. Nine items cannot capture nuance, and the tool does not pretend to. It is a sorting device applied early in an episode, meant to route a decision that would otherwise be made on intuition, and it is at its least useful when someone tries to read fine detail off a single point of difference. It is also a snapshot of one day. A week of bad sleep, a bad shift, or a frightening remark from someone at work can move several of the distress items, and the form has no way of knowing that.

How the three risk groups are worked out

The arithmetic is simple and worth knowing, because the start back risk groups are not evenly spaced bands of a single score. The total score decides whether you are in the low-risk group at all; if you are not, the psychosocial subscale decides whether you are medium or high 1Ref 1Hill JC, Dunn KM, Lewis M, et al. (2008).A primary care back pain screening tool: identifying patient subgroups for initial treatment.The STarT Back Screening Tool's construction and scoring: nine items covering referred leg pain, comorbid pain, two disability items, bothersomeness, catastrophizing, fear, anxiety and depression, with the final five forming a psychosocial subscale; a total range of 0-9 and a subscale range of 0-5; higher scores meaning greater risk of persistent disabling low back pain; the tool being prognostic rather than diagnostic; the three-subgroup cutoffs (0-3 low; >=4 with subscale >=4 high; >=4 with subscale <=3 medium); the development sample of 131 and independent external validation sample of 500 UK primary-care patients; the reported reliability and validity and acceptability; the intended matching of care intensity to subgroup; and the explicit absence of any MCID or MDC for this instrument.. As originally defined:

So the jump from medium to high is not about having more pain. It is about how much of the total is being carried by the five distress items 1Ref 1Hill JC, Dunn KM, Lewis M, et al. (2008).A primary care back pain screening tool: identifying patient subgroups for initial treatment.The STarT Back Screening Tool's construction and scoring: nine items covering referred leg pain, comorbid pain, two disability items, bothersomeness, catastrophizing, fear, anxiety and depression, with the final five forming a psychosocial subscale; a total range of 0-9 and a subscale range of 0-5; higher scores meaning greater risk of persistent disabling low back pain; the tool being prognostic rather than diagnostic; the three-subgroup cutoffs (0-3 low; >=4 with subscale >=4 high; >=4 with subscale <=3 medium); the development sample of 131 and independent external validation sample of 500 UK primary-care patients; the reported reliability and validity and acceptability; the intended matching of care intensity to subgroup; and the explicit absence of any MCID or MDC for this instrument.. Two people can both score 6 out of 9 and land in different groups.

A worked example makes that concrete. Suppose two people each score 6. The first has all four physical items — leg pain, pain elsewhere, and both disability items — plus two of the five distress items. Total 6, subscale 2, medium risk. The second has one physical item and all five distress items. Total 6, subscale 5, high risk 1Ref 1Hill JC, Dunn KM, Lewis M, et al. (2008).A primary care back pain screening tool: identifying patient subgroups for initial treatment.The STarT Back Screening Tool's construction and scoring: nine items covering referred leg pain, comorbid pain, two disability items, bothersomeness, catastrophizing, fear, anxiety and depression, with the final five forming a psychosocial subscale; a total range of 0-9 and a subscale range of 0-5; higher scores meaning greater risk of persistent disabling low back pain; the tool being prognostic rather than diagnostic; the three-subgroup cutoffs (0-3 low; >=4 with subscale >=4 high; >=4 with subscale <=3 medium); the development sample of 131 and independent external validation sample of 500 UK primary-care patients; the reported reliability and validity and acceptability; the intended matching of care intensity to subgroup; and the explicit absence of any MCID or MDC for this instrument.. Identical totals, different pathways, and the second person is the one the tool exists to find, because nothing about their physical picture would have flagged them.

A caution the developers are explicit about: STarT Back reports no minimal clinically important difference and no minimal detectable change 1Ref 1Hill JC, Dunn KM, Lewis M, et al. (2008).A primary care back pain screening tool: identifying patient subgroups for initial treatment.The STarT Back Screening Tool's construction and scoring: nine items covering referred leg pain, comorbid pain, two disability items, bothersomeness, catastrophizing, fear, anxiety and depression, with the final five forming a psychosocial subscale; a total range of 0-9 and a subscale range of 0-5; higher scores meaning greater risk of persistent disabling low back pain; the tool being prognostic rather than diagnostic; the three-subgroup cutoffs (0-3 low; >=4 with subscale >=4 high; >=4 with subscale <=3 medium); the development sample of 131 and independent external validation sample of 500 UK primary-care patients; the reported reliability and validity and acceptability; the intended matching of care intensity to subgroup; and the explicit absence of any MCID or MDC for this instrument.. It is a risk-stratification screener, not an outcome measure, so re-administering it every visit to watch the number fall is using it for something it was not validated to do. If a number is quoted to you as the meaningful change on STarT Back, it is worth asking where it came from, because the development paper does not supply one.

Why a back questionnaire asks about worry and mood

Because the alternative sorting method does not work. The great majority of low back pain is nonspecific low back pain, meaning it cannot be attributed to a specific identifiable pathology, and imaging findings correlate poorly with symptoms 2Ref 2Hartvigsen J, Hancock MJ, Kongsted A, et al. (2018).What low back pain is and why we need to pay attention.That low back pain is the leading cause of years lived with disability worldwide, that most low back pain is non-specific and cannot be attributed to a specific pathology, and that imaging findings correlate poorly with symptoms — the reason structural information cannot be used to sort patients by likely course.. Low back pain is also the leading cause of years lived with disability worldwide, so this is not a small population being sorted 2Ref 2Hartvigsen J, Hancock MJ, Kongsted A, et al. (2018).What low back pain is and why we need to pay attention.That low back pain is the leading cause of years lived with disability worldwide, that most low back pain is non-specific and cannot be attributed to a specific pathology, and that imaging findings correlate poorly with symptoms — the reason structural information cannot be used to sort patients by likely course..

That combination creates the problem STarT Back exists to solve. If most cases have no identifiable structural cause, then structural information cannot tell you which of two people with identical scans will still be off work in six months. Something has to, and the variables that carry the most predictive weight turned out to include how bothersome the pain feels, whether the person catastrophizes about it, whether they are afraid to move, and whether they are anxious or low in mood 1Ref 1Hill JC, Dunn KM, Lewis M, et al. (2008).A primary care back pain screening tool: identifying patient subgroups for initial treatment.The STarT Back Screening Tool's construction and scoring: nine items covering referred leg pain, comorbid pain, two disability items, bothersomeness, catastrophizing, fear, anxiety and depression, with the final five forming a psychosocial subscale; a total range of 0-9 and a subscale range of 0-5; higher scores meaning greater risk of persistent disabling low back pain; the tool being prognostic rather than diagnostic; the three-subgroup cutoffs (0-3 low; >=4 with subscale >=4 high; >=4 with subscale <=3 medium); the development sample of 131 and independent external validation sample of 500 UK primary-care patients; the reported reliability and validity and acceptability; the intended matching of care intensity to subgroup; and the explicit absence of any MCID or MDC for this instrument..

This is regularly misheard as an accusation, so it is worth saying flatly. Scoring high on the psychosocial items does not mean the pain is imagined, exaggerated, or psychological in origin. It means the questionnaire has identified a pattern statistically associated with slower recovery — usually in people whose pain is entirely real and often severe.

The practical inversion is what makes the tool interesting. Distress is not only a consequence of persistent pain; on this evidence it is also one of the better available predictors of it, early in an episode, before anyone can tell from the pain itself 1Ref 1Hill JC, Dunn KM, Lewis M, et al. (2008).A primary care back pain screening tool: identifying patient subgroups for initial treatment.The STarT Back Screening Tool's construction and scoring: nine items covering referred leg pain, comorbid pain, two disability items, bothersomeness, catastrophizing, fear, anxiety and depression, with the final five forming a psychosocial subscale; a total range of 0-9 and a subscale range of 0-5; higher scores meaning greater risk of persistent disabling low back pain; the tool being prognostic rather than diagnostic; the three-subgroup cutoffs (0-3 low; >=4 with subscale >=4 high; >=4 with subscale <=3 medium); the development sample of 131 and independent external validation sample of 500 UK primary-care patients; the reported reliability and validity and acceptability; the intended matching of care intensity to subgroup; and the explicit absence of any MCID or MDC for this instrument..

What the result is supposed to change

The purpose of sorting is matching: giving each group the amount and kind of care it is likely to need, instead of giving everyone the same thing. Low-risk patients are meant to receive reassurance, advice and a short course of the basics. Medium-risk patients are directed toward physiotherapy. High-risk patients are directed toward treatment that addresses the psychosocial obstacles alongside the physical ones 1Ref 1Hill JC, Dunn KM, Lewis M, et al. (2008).A primary care back pain screening tool: identifying patient subgroups for initial treatment.The STarT Back Screening Tool's construction and scoring: nine items covering referred leg pain, comorbid pain, two disability items, bothersomeness, catastrophizing, fear, anxiety and depression, with the final five forming a psychosocial subscale; a total range of 0-9 and a subscale range of 0-5; higher scores meaning greater risk of persistent disabling low back pain; the tool being prognostic rather than diagnostic; the three-subgroup cutoffs (0-3 low; >=4 with subscale >=4 high; >=4 with subscale <=3 medium); the development sample of 131 and independent external validation sample of 500 UK primary-care patients; the reported reliability and validity and acceptability; the intended matching of care intensity to subgroup; and the explicit absence of any MCID or MDC for this instrument..

The argument for bothering is that undifferentiated care goes wrong in both directions. Low-value care for low back pain — unnecessary imaging, opioids, injections and surgery — is widespread internationally and is a target for reduction, which is the over-treatment side 5Ref 5Buchbinder R, van Tulder M, Öberg B, et al. (2018).Low back pain: a call for action.That low-value care for low back pain — unnecessary imaging, opioids, injections and surgery — is widespread globally and should be reduced, which is the over-treatment half of the argument for matching care intensity to risk.. The under-treatment side is the person who needed structured help early, got a leaflet, and is still in trouble a year later.

What sits underneath the medium and high pathways is exercise-based care, and there the evidence is reasonably settled: exercise therapy probably reduces pain and improves function in chronic non-specific low back pain compared with no treatment, usual care or placebo, with small-to-moderate effects 6Ref 6Hayden JA, Ellis J, Ogilvie R, Malmivaara A, van Tulder MW (2021).Exercise therapy for chronic low back pain.That exercise therapy probably reduces pain and improves function in chronic non-specific low back pain compared with no treatment, usual care or placebo, with small-to-moderate effects — the evidence underneath the medium- and high-risk care pathways.. Small-to-moderate is an honest description rather than a disappointing one — it is roughly what most single treatments achieve in this condition.

One limit has to be stated plainly. The 2008 development paper establishes that the tool sorts people, and how well it does that 1Ref 1Hill JC, Dunn KM, Lewis M, et al. (2008).A primary care back pain screening tool: identifying patient subgroups for initial treatment.The STarT Back Screening Tool's construction and scoring: nine items covering referred leg pain, comorbid pain, two disability items, bothersomeness, catastrophizing, fear, anxiety and depression, with the final five forming a psychosocial subscale; a total range of 0-9 and a subscale range of 0-5; higher scores meaning greater risk of persistent disabling low back pain; the tool being prognostic rather than diagnostic; the three-subgroup cutoffs (0-3 low; >=4 with subscale >=4 high; >=4 with subscale <=3 medium); the development sample of 131 and independent external validation sample of 500 UK primary-care patients; the reported reliability and validity and acceptability; the intended matching of care intensity to subgroup; and the explicit absence of any MCID or MDC for this instrument.. Whether care organised around the sorting produces better outcomes than care that ignores it is a separate question, answered by separate trials, and it should not be attributed to this paper. A clinician using STarT Back is using a validated way of grouping patients, which is not the same claim as a validated way of treating them.

Where STarT Back sits among the back-pain questionnaires

Someone with low back pain may be handed three or four forms in a single appointment, and they are not interchangeable. Sorting them by the question each one answers makes the pile much easier to read, and makes it obvious when a number is being used for something it cannot support.

• How likely is this to persist? STarT Back, nine items, a total out of nine and a psychosocial subscale out of five, producing low, medium or high risk 1Ref 1Hill JC, Dunn KM, Lewis M, et al. (2008).A primary care back pain screening tool: identifying patient subgroups for initial treatment.The STarT Back Screening Tool's construction and scoring: nine items covering referred leg pain, comorbid pain, two disability items, bothersomeness, catastrophizing, fear, anxiety and depression, with the final five forming a psychosocial subscale; a total range of 0-9 and a subscale range of 0-5; higher scores meaning greater risk of persistent disabling low back pain; the tool being prognostic rather than diagnostic; the three-subgroup cutoffs (0-3 low; >=4 with subscale >=4 high; >=4 with subscale <=3 medium); the development sample of 131 and independent external validation sample of 500 UK primary-care patients; the reported reliability and validity and acceptability; the intended matching of care intensity to subgroup; and the explicit absence of any MCID or MDC for this instrument..
• How disabled am I right now? The Oswestry Disability Index, ten fixed sections reported as a percentage 4Ref 4Fairbank JCT, Pynsent PB (2000).The Oswestry Disability Index.That the Oswestry Disability Index is a validated ten-section patient-reported measure of current low-back-pain-related disability scored as a percentage — used here as the contrast case for a severity measure rather than a prognostic one.. The quebec back pain scale answers the same kind of question with a different item set, as do several other disability indices.
• Is anything dangerous going on? Not a questionnaire at all — a history and examination looking for red flags, interpreted with full knowledge of how often individual red flags raise a false alarm 3Ref 3Downie A, Williams CM, Henschke N, et al. (2013).Red flags to screen for malignancy and fracture in patients with low back pain: systematic review.That screening for serious underlying causes of low back pain is done with red flags rather than a risk score, and that most individual red flags carry high false-positive rates while some — older age, prolonged corticosteroid use, and significant trauma when screening for fracture — meaningfully raise post-test probability..
• Do I need a scan? A separate decision again, and one where the default in early, uncomplicated cases is no. This is the territory of the mri for back pain conversation rather than of any risk score.

Read as a set, they describe different axes of the same episode, and the most common error is collapsing them. A high STarT Back score is not a severe injury, a low one is not a promise, and neither of them says anything about whether the pain in your back has a dangerous cause. That last question is answered by a clinician who can examine you, and it is the one worth raising first if anything about the episode feels unlike ordinary back pain.