Classification of Low Back Pain in Primary Care:
Using "Bothersomeness" to Identify
the Most Severe CasesThis section is compiled by Frank M. Painter, D.C.
Send all comments or additions to: Frankp@chiro.org
FROM: Spine (Phila Pa 1976) 2005 (Aug 15); 30 (16): 1887–1892 ~ FULL TEXT
Kate M Dunn, Peter R Croft
Primary Care Sciences Research Centre,
Keele University,
Staffordshire, ST5 5BG,
United Kingdom.
k.m.dunn@cphc.keele.ac.uk
Study design: Prospective inception cohort.
Objective: To investigate the validity and use of a single question on the bothersomeness of low back pain (LBP) as a method of classifying the severity of symptoms among patients seen in clinical practice.
Summary of background data: There is no widely accepted method for classifying patients with nonspecific LBP in clinical practice. There have been no previous reports of the validity and use of a question on bothersomeness as a method of classifying patients with LBP.
Methods: Consecutive patients (30–59 years old) with LBP consulting at 5 United Kingdom general practices (n = 1,464) were mailed a questionnaire after consultation and 6 months later. Construct validity was assessed by comparing baseline responses on the single bothersomeness question with pain, disability, general health, and psychologic health measures. The ability of bothersomeness to predict outcomes at 6 months was assessed against pain, disability, work absence, and health care use.
Results: A total of 935 patients (65%) completed baseline questionnaires. There were 776 (83%) patients who consented to enter the follow-up study, and 447 returned the 6–month questionnaire. At baseline, bothersomeness correlated with pain, disability, and other measures (P < 0.001), and had 80% sensitivity (61% specificity) to detect people in the highest category of pain and disability. People with bothersome LBP at baseline evaluation had an increased risk of work absence or health care consultations for LBP 6 months later (relative risks 2.8 and 1.9, respectively).
Conclusions: There is evidence for the validity of a single bothersomeness question as a measure of LBP severity. It has the potential to provide a practical standard scheme for classifying patients with LBP in clinical practice. However, further work is needed to clarify its usefulness in a clinical setting.
Key words : low back pain, primary health care, health surveys, classification, sensitivity and specificity.
From the FULL TEXT Article:
Background
Low back pain (LBP) is among the most common reasons for a primary care consultation in the United Kingdom, [1] and most patients with LBP seeking health care will first consult their general practitioner. Diagnostic triage for “red flags” in LBP is well documented. [2] However, once specific serious causes have been excluded, there is no evidence-based agreement regarding classification and prioritization of nonspecific LBP among primary care patients, and alternative methods of classifying patients, which are not based on etiology, are needed. One such method would be to identify patients based on subjective or objective measures of severity, which are associated with outcomes. However, classification of patients with LBP is notoriously difficult because of the variable and recurrent course of pain and disability. Von Korff et al [3] suggested a grading system for use in population-based and primary care surveys, which classified chronic patients with LBP into 1 of 5 grades of severity using a 7–item Guttman scale. This system has been used in a variety of settings, and found to be a valid and reliable method of classification in surveys. [3–6] Although this may be seen as a form of “reference standard,” it was not designed for use in clinical settings, and the administration and scoring is likely to be too long and complicated for practical use during primary care consultations.
A core set of LBP outcome measures was suggested by a multinational group of investigators, [7] including questions referring to bothersomeness, although it was pointed out that bothersomeness, unlike the other measures, had not been used or assessed outside the United States. Since this publication, bothersomeness has been assessed in an Israeli study of LBP, [8] but not elsewhere, as far as the authors are aware. The use of bothersomeness was first proposed in 1994 [9] for assessing symptom severity among patients with asthma. It was chosen from a number of questions after piloting because it was acceptable to both patients and clinicians. Its conceptual basis lay in the need to find a single term that could function in clinical situations as a simple summary of outcome for specific symptoms or diseases. Following this, bothersomeness was reported to be extremely important for assessing urinary tract symptoms [10] and was also used among patients with sciatica. [11] Bothersomeness can be assessed using 1 question with 5 simple response categories and would seem likely to be practical for use as part of a primary care LBP consultation.
The purpose of this study was 2–fold. The first purpose was to investigate the validity of a simple question on bothersomeness against established measures of severity and indicators of outcome in a sample of primary care patients with LBP in the United Kingdom. The second purpose was to determine the potential usefulness of this question as a method to classify primary care patients with LBP in clinical practice.
Materials and Methods
This work is part of the Backpain Research in North Staffordshire Study. This is a longitudinal study looking at pain, disability, and health care use in a consecutive cohort of patients with LBP consulting in primary care, using epidemiologic and qualitative methods. The North Staffordshire Local Research Ethics Committee approved this study.
Setting. Five computerized general practices in North Staffordshire that are part of the North Staffordshire General Practice Research Network took part in the study. They cover a heterogeneous population both socioeconomically and geographically. All patients requiring care for their LBP through the National Health Service will first consult a primary care general practitioner, who will then refer them to more specialist services such as physiotherapy or surgery if appropriate. In the United Kingdom, approximately 98% of the population are registered with a National Health Service general practitioner. [12]
Recruitment. Contact information for all patients aged 30–59 years consulting their general practitioner with LBP during the study (mid October 2001 to mid October 2002) was downloaded each week from the practice databases. These patients were sent a letter from their general practice along with an information sheet and questionnaire, inviting them to participate in the study by returning the questionnaire in a prepaid envelope to the research unit at Keele University. Patients consulting more than once during the study were only invited to participate in the study after their first consultation. Most patients were mailed information during the week following consultation, and the maximum time between consultation and mailing was 3 weeks. For nonresponders, a reminder postcard was sent after 1 week and a reminder questionnaire after 3 weeks. The last page of the questionnaire contained a form asking patients whether they gave their consent to be followed up as part of the study. These signed forms were detached from the questionnaire on return to maintain confidentiality.
Computerized primary care records in the United Kingdom are recorded using the Read Code classification system, and patients were identified through the use of Read Codes indicating a LBP consultation. A wide range of codes were used to make the selection as sensitive as possible because most patients with LBP cannot be given a specific diagnosis when seen in primary care. [13] The codes selected were intended to include all cases of LBP, and in doing so, may include a small number that could relate to thoracic back pain. Codes indicating a red flag diagnosis (e.g., cauda equina syndrome, significant trauma, ankylosing spondylitis, cancers) or referring only to the cervical spine, hip, congenital disorders, or to adolescents, were excluded. The validity of Read Codes in electronic patient records in the United Kingdom has been established in previous research. [14] In addition, the practices involved in the study regularly have their coding practices audited as part of their involvement in the North Staffordshire General Practice Research Network.
Data Collection. All patients were sent postal self-completion questionnaires at baseline and 6 months, including the following instruments. The Roland-Morris Disability Questionnaire (RMDQ) [15] measures self-reported disability from back pain and asks patients to think of themselves “today.” Scores from the modified version used by Patrick et al [11] ranging from zero (no disability) to 23 (highest disability) are reported here. The definition of a high level of dysfunction was a modified RMDQ score >14 as used by Cherkin et al. [16] The Chronic Pain Grade (CPG)3 classifies individuals into 1 of 5 grades of chronic pain, specifically chronic back pain in this study. Of the 7 questions, 6 relate to the previous 6 months, the other asks about pain at the present time. The grades produced are: zero, pain free; I, low disability, low intensity; II, low disability, high intensity; III, high disability, moderately limiting; and IV, high disability, severely limiting. The Short-Form 36 questionnaire, version 2 (SF-36v2TM) [17] is a commonly used health status instrument and comprises 8 dimensions, each scored from zero (worst) to 100 (best health status).
The Hospital Anxiety and Depression Scale [18] is used to assess aspects of psychologic status, and provides scores for anxiety and depression from zero (lowest) to 21 (highest psychologic distress) for each. Questions relate to the previous 2 weeks. Probable cases of clinical anxiety or depression were defined as scores of 11 or more. [19] The reliability of these 4 measures has been established. [3, 5, 6, 8, 11, 20, 21] Patients were also asked whether they were currently absent from work as a result of their LBP (either employed or unemployed), and whether they had consulted a health care professional (e.g., general practitioner, hospital physician or surgeon, physiotherapist, osteopath or chiropractor) about their LBP during the previous 4 weeks.
Pain intensity was estimated through the mean of 4, 11– point numerical rating scales for pain “right now,” and worst, least and average pain (as recommended by Jensen et al [22]) relating to the last 2 weeks. Patients were also asked if their pain had spread down their legs or below their knee (distal leg pain) during the last 2 weeks. Patients were asked about the bothersomeness of their back pain [7] with the question “In the last 2 weeks, how bothersome has your back pain been?” There were 5 possible responses: “not at all,” “slightly,” “moderately,” “very much,” and “extremely.”
The duration of symptoms was determined through recall of the last pain-free month, which was based on the definition proposed by de Vet et al. [23] General demographic questions were also included. Questions were ordered chronologically according to the period they referred to because this may elicit higher and quicker response. [5]
Analysis. Responses were logged, and data were entered into Microsoft Access 97 software (Microsoft, Corp., Redmond, WA). Analysis was performed using SPSS (SPSS, Inc., Chicago, IL) for Windows. [24] Scores for each questionnaire were calculated according to the methods specified by the questionnaire developers. Associations among bothersomeness and other variables were assessed using analysis of variance or χ2 tests for trend. Bothersomeness was dichotomized (unless otherwise specified) into bothersome (i.e., reports of “very much” or “extremely” bothersome back pain) and not bothersome (i.e., reports of “moderately,” “slightly,” or “not at all” bothersome back pain). At baseline, bothersomeness and other individual questions were examined against a reference standard of Chronic Pain Grade IV (CPG IV), by calculation of sensitivity, specificity, and the area under the receiver operating characteristic (ROC) curve. The ability of baseline bothersomeness to detect people with CPG IV, absent from work as a result of LBP, or consulting a health care professional about LBP at 6 months was assessed using relative risks.
Results
During the 1–year study, 1,464 patients consulted with LBP at the 5 study general practices. This total represents 9.3% of the practice population, aged 30–59 years, with specific annual consultation prevalence figures ranging from 6% of males aged 30–34 years to 12% of females aged 40–44 years. Approximately half the patients were male (47%), and mean age was 44.7 years in both males and females.
Response
Baseline questionnaires were returned by 935 patients with LBP. Seven questionnaires were returned with addressee unknown, 1 patient died, 1 had severe learning difficulties and was unable to complete the questionnaire (notification by career), and 7 telephoned to say they had not consulted their general practitioner with LBP, giving an adjusted response to baseline of 65%. The approximate median time between primary care consultation and return of questionnaire was 17 days (the precise date of consultation was unavailable).
The mean age of baseline responders was 45.6 years (standard deviation [SD] 8.45), approximately the same as the total group of patients consulting. Females were slightly more likely than males to return questionnaires. Just less than half the baseline responders (45%, n = 416) was educated after age 16 years. Two-thirds (65%, n = 597) of the responders were employed at the time of completing the questionnaire. Of baseline responders, 82% had back pain for more than a day before the current episode. Of responders, 22% could be classified as having acute pain <3 months since their last back painfree month), but for more than 50% of respondents, it was more than a year since they had last had a whole month in which they were free of back pain.
Of the 935 baseline responders, 776 people (83%) agreed to be followed up, and 447 (60%) of these returned the 6–month questionnaire. People agreeing to be followed up (consenters) had the same age and gender characteristics as nonconsenters, but consenters had slightly higher levels of pain and disability, and poorer psychologic status on the baseline questionnaire than nonconsenters. There were no differences between responders and nonresponders to the 6–month questionnaire in terms of pain, disability, or psychologic status at baseline.
Associations with Bothersomeness at Baseline
The distributions of the CPG and the bothersomeness question are presented in Table 1. This result indicates that 31% of people were classified with CPG IV (i.e., high disability and severely limiting LBP), and 52% considered their back pain to be very much or extremely bothersome. Of the total sample, 24% had both CPG IV and had very or extremely bothersome back pain. Bothersomeness showed associations with measures of pain and disability, including the CPG, and with psychologic health (Tables 1, 2), all of which were statistically significant (P <0.001). This result can be seen as evidence of the construct validity of bothersomeness in this United Kingdom primary care LBP population.
Bothersomeness as a Screening Tool at Baseline
Figure 1 presents the sensitivity and specificity of various cutoffs for dichotomizing the bothersomeness question assessed against the reference standard of CPG IV at baseline. This is shown as a ROC curve of sensitivity against one minus specificity; the area under the curve is 0.75. Taking the point nearest the top left of the chart as the cutoff, this clearly shows that defining “bothersome” LBP as a combination of “extremely bothersome” and “very much bothersome” is the most appropriate. This point on a ROC curve is usually pe rceived to be the most appropriate where reasonable levels of both sensitivity and specificity are needed. This definition of bothersomeness gives a sensitivity of 80% (95% confidence interval [CI] 75% to 84%) and specificity of 61% (95% CI 57% to 65%) to detect patients with CPG IV. These sensitivity and specificity figures are altered little when the data are stratified by gender, age group, duration of symptoms, or presence or absence of leg pain or distal leg pain.
A sensitivity of 80% and specificity of 61% are reasonable for a single question, but the addition of other simple questions may improve its use as a screening tool. Figure 1 presents the sensitivity and specificity for a number of different questions and combinations of questions. This shows that the use of specific questions regarding disability in addition to bothersomeness can improve specificity to detect patients with LBP in CPG IV. The addition of questions on whether patients had to walk more slowly or avoid heavy jobs because of their back pain increased specificity to 68%, while sensitivity was maintained at 79%. Questions regarding only being able to walk short distances or doing less daily work increased specificity to 76%, but sensitivity decreased to 73%. The addition of other combinations of questions regarding doing less daily work, avoiding heavy jobs around the house, only able to stand for short periods, walking more slowly, or walking short distances gave similar results. Other questions and combinations of questions either did not improve specificity or only improved specificity at a significant loss of sensitivity.
6–Month Follow-up
The analysis in Table 2was repeated using bothersomeness at baseline and measures of pain, disability, and psychologic health at 6 months. Associations of a similar size to those observed at baseline were found in this prospective analysis, and all were statistically significant (P<0.01). For example, mean 6–month modified RMDQ scores ranged from 2.2 (SD 4.1) for people who said their back pain at baseline was not at all bothersome, to 10.7 (SD 7.6) for people who had extremely bothersome back pain at baseline (other data not shown).
Changes in ratings of bothersomeness were related to changes in other study variables over the 6–month period. People who had improved ratings of bothersomeness at 6 months (n = 231) also had improved pain and disability (mean improvements 3.0 [SD 2.0] in pain intensity and 5.2 [SD 5.3] on the modified RMDQ). Similarly, people whose bothersomeness did not change (n = 164) also had little change on other measures (mean improvements 0.3 [SD 1.1] in pain intensity and 0.12 [SD 3.7] on the modified RMDQ). Finally, people whose bothersomeness had deteriorated (n = 47) had worse pain and disability scores at 6 months than baseline (mean deterioration 1.5 [SD 1.7] in pain intensity and 1.7 [SD 4.1] on the modified RMDQ).
Patients with bothersome LBP at baseline had 3 times the risk of having CPG IV 6 months later compared with patients not reporting bothersome LBP at baseline (relative risk 3.13, 95% CI 2.06, 4.77). Similarly, patients with bothersome LBP at baseline had approximately 3 times the risk of being absent from work as a result of LBP (relative risk 2.79, 95% CI 1.72, 4.51), and almost double the risk of still consulting health care professionals about their LBP (relative risk 1.87, 95% CI 1.43, 2.44) at 6 months, compared to people without bothersome LBP at baseline.
Discussion
This study has illustrated the validity of bothersomeness as a measure of severity in a group of primary care patients with LBP in the United Kingdom. Associations between a single question on bothersomeness have been shown with measures of pain, disability, psychologic health, and work absence. The ability of the question to detect the LBP patients with the highest pain and disability has also been shown, and it has been shown to be a predictor of outcome 6 months later. In the United States, bothersomeness has previously been used as a measure of severity among primary care patients with LBP, [16, 25] and has been shown to have associations with functional status, work absence, and other LBP factors in patients with sciatica. [11, 26] Bothersomeness has also been found to be valid and reliable in Israeli patients with LBP. [8]
A single question on bothersomeness identified 80% of severely disabled patients with LBP among a cohort of LBP primary care patients. The addition of questions on limitation of daily activities from a standard LBP disability scale15 improved the specificity for identification of severe cases. Given the large numbers of patients consulting for LBP in primary care (9.3% of adults aged 30–59 years in this study), a simple method for classifying cases in a standard way is likely to be useful where time for form filling and resources for treatment are limited. Furthermore, a recent study found that most primary care providers did not inquire about functional limitations during LBP consultations, [13] and a simple question about bothersomeness might be an appropriate way to change this culture and improve patient communication.
Simple questions on bothersomeness and disability may also be practical and useful in a clinical primary care setting for categorizing patients in terms of their severity, and to provide a standard measure for monitoring outcomes across a range of practices and practitioners. Bothersomeness is widely recognized as a useful tool for classifying patients with urinary tract symptoms in primary care, [10, 27] and is used as an indicator of severity in asthma [28] and migraine. [29] This work indicates that bothersomeness could also be useful for LBP in primary care as a standard classification for choosing and evaluating treatment or further investigation.
The category “very much” or “extremely” bothersome in response to the single question on bothersomeness has a sensitivity of 80% for severe pain and disability according to the CPG. This is good for a screening tool, but, inevitably, some people are misclassified. Analysis of the people who were “misclassified” according to the reference standard of the CPG indicates that the people with bothersome LBP but not CPG IV are likely to have LBP that has a higher impact on their lives because they report higher pain and disability and poorer psychologic health, than people classified with CPG IV but who did not report bothersome LBP. Therefore, bothersomeness may actually be a more appropriate measure of severity in primary care patients with LBP than the CPG because it is selecting people with potential psychosocial barriers to recovery.
There are some limitations to this study. First, although the population studied was primary care patients with LBP with similar characteristics to a comparable study in the United States, [3] which adds strength to its likely applicability to clinical settings, the results are based on responses to a questionnaire mailed up to a week after consultation. It should be emphasized that the time of consultation and the point at which baseline measurements were reported are not identical. Further investigation of the practicality and usefulness of these questions in a clinical setting is necessary. In addition, some study participants may have been misclassified with LBP and may actually have pain in other regions, such as thoracic back pain; the number of such patients is likely to be small. Second, because there was incomplete follow- up, nonrespondents could represent a different group to respondents, although we have given data to show that respondents to follow-up were very similar to all baseline respondents.
Any potential bias from baseline nonresponse could plausibly affect our estimate of the prevalence of bothersomeness, but it is unlikely that the associations with bothersomeness reported here would be influenced. Third, the calculation of sensitivity and specificity was based on a reference standard of the CPG,3 which has been assessed for validity and reliability in surveys,3–6 and is widely accepted, but it may not be ideal for identifying severe LBP cases in primary care. However, in the absence of a widely used or validated clinical classification of patients with nonspecific LBP, this seems acceptable as a reference standard measure of severity. In addition, in the analysis of prospective data, bothersomeness has been a predictor of work absence and health care use, as well as the CPG.
This study has investigated the use of a single question on bothersomeness in primary care patients with LBP, and provided evidence of its construct and predictive validity as a measure of severity. Its potential for use in clinical practice has been established as a method to identify in a standard way the most severely affected patients and predict outcome 6 months later. However, its usefulness and practicality in a clinical setting has not been studied, and further work is needed to clarify this. In conclusion, a simple question on bothersomeness could be useful for classifying patients with LBP at primary care consultation, and for standardized reporting and audit in clinical practice.
Key Points
There is no universally accepted classification scheme for patients with LBP in clinical practice.
This study provides evidence for the validity of a single question on bothersomeness as a measure of LBP severity.
People with bothersome LBP have an increased risk of being absent from work or consulting health care professionals for their LBP 6 months later.
Bothersomeness has the potential to provide a practical, useful, and simple standard scheme for classifying patients with LBP in clinical practice.
References:
Royal College of General Practitioners and Office of Population Censuses and Surveys,
Department of Health. Morbidity Statistics from General Practice, 4th National Survey 1991–1992.
London, UK: HMSO; 1995.Royal College of General Practitioners.
Clinical Guidelines for the Management of Acute Low Back Pain
London: Royal College of General Practitioners; (1999). [UK]Von Korff M, Ormel J, Keefe FJ, et al.
Grading the severity of chronic pain.
Pain 1992;50:133–49.Cote P, Cassidy JD, Carroll L.
The Saskatchewan Health and Back Pain Survey.
The Prevalence of Neck Pain and Related Disability in Saskatchewan Adults
Spine (Phila Pa 1976). 1998 (Aug 1); 23 (15): 1689–1698Dunn KM, Jordan K, Croft PR.
Does questionnaire structure influence response in postal surveys?
J Clin Epidemiol 2003;56:10–6.Smith BH, Penny KI, Purves AM, et al.
The Chronic Pain Grade questionnaire: Validation and reliability in postal research.
Pain 1997;71:141–7.Deyo RA, Battie M, Beurskens AJ, et al.
Outcome measures for low back pain research. A proposal for standardized use.
Spine 1998;23:2003–13.Jacob T, Baras M, Zeev A, et al.
Low back pain: Reliability of a set of pain measurement tools.
Arch Phys Med Rehabil 2001;82:735–42.Steen N, Hutchinson A, McColl E, et al.
Development of a symptom based outcome measure for asthma.
BMJ 1994;309:1065–8.Abrams P.
Managing lower urinary tract symptoms in older men.
BMJ 1995; 310:1113–7.Patrick DL, Deyo RA, Atlas SJ, et al.
Assessing health-related quality of life in patients with sciatica.
Spine 1995;20:1899–908.Bowling A, Bond M, Jenkinson C, et al.
Short Form 36 (SF-36) Health Survey questionnaire: Which normative data
should be used? Comparisons between the norms provided by the Omnibus Survey
in Britain, the Health Survey for England and the Oxford Healthy Life Survey.
J Public Health Med 1999;21: 255–70.Turner JA, Leresche L, Von KorffM, et al.
Back pain in primary care: Patient characteristics, content of initial visit,
and short-term outcomes.
Spine 1998; 23:463–9.Hassey A, Gerrett D, Wilson A.
A survey of validity and utility of electronic patient records in a general practice.
BMJ 2001;322:1401–5.Roland M, Morris R.
A study of the natural history of back pain. Part I: Development of a reliable
and sensitive measure of disability in low-back pain.
Spine 1983;8:141–4.Cherkin DC, Wheeler KJ, Barlow W, et al.
Medication use for low back pain in primary care.
Spine 1998;23:607–14.Ware JE Jr.
SF-36 Health Survey Update.
Spine 2000;25:3130–9.Zigmond AS, Snaith RP.
The Hospital Anxiety and Depression Scale.
Acta Psychiatr Scand 1983;67:361–70.Dowell AC, Biran LA.
Problems in using the hospital anxiety and depression scale for screening patients in general practice.
Br J Gen Pract 1990;40:27–8.Jenkinson C, Stewart-Brown S, Petersen S, et al.
Assessment of the SF-36 version 2 in the United Kingdom.
J Epidemiol Community Health 1999;53: 46–50.Herrmann C.
International experiences with the hospital anxiety and depression scale—
A review of validation data and clinical results.
J Psychosom Res 1997;42:17–41.Jensen MP, Turner LR, Turner JA, et al.
The use of multiple-item scales for pain intensity measurement in chronic pain patients.
Pain 1996;67:35–40.de Vet HCW, Heymans MW, Dunn KM, et al.
Episodes of low back pain: A proposal for uniform definitions to be used in research.
Spine 2002;27: 2409–16.SPSS for Windows [computer program].
Rel. 10.0.7. Chicago, IL: SPSS Inc., 2000.Cherkin DC, Deyo RA, Street JH, et al.
Predicting poor outcomes for back pain seen in primary care using patients’ own criteria.
Spine 1996;21: 2900–7.Atlas SJ, Deyo RA, Keller RB, et al.
The Main Lumbar Spine Study, Part II. 1-year outcomes of surgical
and nonsurgical management of sciatica.
Spine 1996;21:1777–86.Peters TJ, Donovan JL, Kay HE, et al.
The International Continence Society “Benign Prostatic Hyperplasia” Study:
The bothersomeness of urinary symptoms.
J Urol 1997;157:885–9.Althuis MD, Sexton M, Prybylski D.
Cigarette smoking and asthma symptom severity among adult asthmatics.
J Asthma 1999;36:257–64.Cramer JA, Silberstein SD, Winner P.
Development and validation of the Headache Needs Assessment (HANA) survey.
Headache 2001;41:402–9.
Return LOW BACK PAIN
Since 5-05-2021
| Home Page | Visit Our Sponsors | Become a Sponsor |
Please read our DISCLAIMER |