Rheumatology 2001; 40: 528-536
© 2001 British Society for Rheumatology
Original Papers |
A brief screening tool for knee pain in primary care. 1. Validity and reliability
Primary Care Sciences Research Centre and
1 School of Health, Keele University, Keele, Staffordshire ST5 5BG, UK
 |
Abstract |
|---|
 Top Abstract IntroductionMethodResultsDiscussionReferences |
|---|
Objectives. To design and test the performance of a new knee pain screening tool (KNEST), both separately and together with a combination of existing questionnaires, which will be used to assess the general health status of knee pain sufferers in primary care.
Methods. A postal survey of knee pain and disability was sent to a random sample of 240 individuals aged over 55 yr registered with two general practices in North Staffordshire. The survey questionnaire consisted of the KNEST; a pain manikin; the Short Form 36 (SF-36); the Hospital Anxiety and Depression Scale (HADS); demographic questions; and the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) for those who reported knee pain. A second, identical questionnaire was sent 2 weeks later to a random subsample of responders (n=80) to test repeatability.
Results. An 85% baseline response rate was achieved for the first questionnaire. The 12-month prevalence of knee pain identified from baseline responders to the survey was 45%. A response rate of 74% was achieved for the repeatability questionnaire. Each section of the questionnaire was well completed and repeatability was good for nearly all measures (most reliability scores exceeded 0.6). A new core question about knee pain showed good internal reliability, with an agreement score of 91% between baseline and retest assessment, and good construct validity in relation to knee pain identified on the pain manikin (agreement 95%). Good agreement was found between recalled consultation for knee pain in the questionnaire and evidence of consultation for knee pain in general practice records.
Conclusions. The KNEST appears to be a reliable and valid composite tool for the study of population needs and outcomes of care for people aged over 55 yr with knee pain.
KEY WORDS: Knee Pain Screening Tool (KNEST), SF-36, WOMAC, HADS, Knee pain, Knee osteoarthritis, Primary care, Needs assessment, GP record review, Epidemiology.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionReferences |
|---|
Osteoarthritis (OA) is the most common cause of pain and disability in older people in the UK and elsewhere [1–4], and the knee is the most common site affected. Traditionally, OA has been diagnosed by radiographic features, including evidence of joint space narrowing, subchondral sclerosis and osteophyte formation. However, it can also be regarded as a clinical syndrome of pain and restricted joint movement. The first presenting symptom of knee OA is often pain in the joint, and in patients over the age of 55 yr most knee pain is attributed to OA. Evidence from morbidity surveys in British primary care confirms that OA and joint pain contribute significantly to the workload of general practice [5].
Good health-care provision for knee OA in primary and secondary care will depend on identifying effective interventions for all grades of the condition and determining the need for those interventions. This paper is concerned with the design of a core set of questions about knee pain for use in the older population, which are directed towards the need for and the provision of health-care in primary care settings. The specific aims of this pilot study were to design and validate this new screening tool, which covers pain and the use of health-care, and to assess its performance when administered with a combination of existing scales for measuring the general health status of older knee pain sufferers.
|
Method |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionReferences |
|---|
Design of the Knee Pain Screening Tool (KNEST)
A core set of questions about knee pain was designed. These are shown in Appendix I and include a new question to act as a sift to identify knee pain sufferers. This new question asks ‘Have you had pain in the last year in or around the knee?’ This question is different from that used in previous studies [6, 7], which asked ‘Have you ever had pain in or around the knee on most days for at least a month?’ and ‘If so, have you experienced any pain during the last year?’
The question about chronicity attempts to capture total days in pain over a recalled period of 12 months, irrespective of whether these were consecutive or not. This method of summarizing chronic pain that is intermittent as well as persistent was derived from a question developed by von Korff et al. for the study of lower back pain [8].
Format of the survey questionnaire
A blank manikin for shading pain and a number of previously well-established questionnaires were collated with the KNEST to form a survey questionnaire instrument. The reason for including other measures was to capture information on a wider range of patient characteristics than can be provided by the KNEST alone, for example general health, anxiety and depression, and disability. These wider patient characteristics are important for the identification of comorbidities and other pain experiences, and are relevant to a more comprehensive assessment of the health-care needs of a population. They will be fully reported in future publications.
The composition of the survey questionnaire is outlined in Table 1
. Consideration was given to the order in which each measure should appear. The first section (Section A) of the questionnaire was the pain manikin (Fig. 1), which enquired about body pain in the past month with the purpose of identifying musculoskeletal comorbidity in relation to knee pain. Previous studies have investigated body charts for back and widespread pain [9–11] but their reliability for scoring knee pain has not been reported hitherto. The body chart does not form a part of the KNEST, but shading in any one of the four knee areas on the chart was used to validate the new knee pain question in the KNEST.
|
|
The second section (Section B) was the Short Form 36 (SF-36). This has been previously validated for use as a generic health-related quality of life measure in general populations [12, 13], and also in selected populations with knee OA [14]. The third section (Section C) consisted of the Hospital Anxiety and Depression Scale (HADS). This is a well-validated measure of psychological well-being [15]. The fourth section (Section D) consisted of demographic questions and two questions from the KNEST relating to pain and injury. Those who reported having had knee pain according to the KNEST were then asked to complete the last section (Section E), which consisted of two parts. Part One included the remaining KNEST items. Part Two was the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC). The WOMAC is a validated disease-specific outcome measure for use in assessing pain, stiffness and physical difficulty in people with hip and knee OA [16]. The WOMAC asks the respondent to enter the amount of pain experienced in the last 48 h because of knee OA. For the purposes of this study the word ‘osteoarthritis’ was not used. Instead, respondents were asked to think about ‘the amount of pain experienced in the last 48 h in the knee’. This minor change made the questionnaire relevant to people who had not yet consulted the general practitioner (GP) or had a diagnosis of OA.
Most of the individual scales in the composite questionnaire (SF-36, HADS and WOMAC) have been validated previously and subjected to rigorous testing [12–20]. The main aim of this study was not, therefore, to test the validity and reliability of these scales. Rather, the aims were (i) to test the performance and reliability of individual items of the new screening tool (the KNEST questions); (ii) to test the validity of the new pain question in the KNEST against the manikin; (iii) to test the validity of the consultation question in the KNEST against records held by the general practice; and (iv) to assess the reliability of the combination of measures used in the survey of which the KNEST formed one part. The local ethics committee agreed the study.
The study had three parts: a prepilot exercise in a sample of out-patients; a formal pilot and repeatability study in a population sample; and a review of general practice records.
Part one: prepilot exercise
Discussions about individual items on the KNEST and the combination of other measures were held with health professionals locally and nationally. The overall questionnaire was then prepiloted on a sample of patients aged over 55 yr attending a local orthopaedic out-patient clinic with knee pain. All patients found the questions on both the KNEST and other measures easy to understand and to complete, and to be relevant and comprehensive. Nearly all patients found the WOMAC instruction page to be confusing. This page of instructions was not, therefore, used in the full pilot study. However, the instructions at the beginning of each section in the WOMAC (Pain, Stiffness and Physical Functioning) were retained. The whole questionnaire took no longer than 25 min to complete.
Part Two: administration of the survey instrument
The names of all people aged over the age of 55 yr registered with two general practices in North Staffordshire were extracted from the GP computer systems. A random sample of 240 subjects (120 from each list) was selected for this study.
Each person was sent a patient information sheet, a covering letter signed by his or her GP, a questionnaire and a prepaid return envelope. After 14 days, non-responders were sent a reminder postcard, and for the purposes of repeatability analysis a random subsample of 80 first-time responders were sent a retest questionnaire. After an additional 10 days the remaining non-responders were sent another baseline questionnaire.
Response rates to the baseline and retest questionnaires and completion rates for the individual items/ scales were recorded. Summary statistics were analysed for individual items of the KNEST, overall and by age and gender.
Reliability and validity were measured in three ways. First, a test–retest analysis (intra-item reliability) was performed to assess the repeatability of questionnaire items. Secondly, an inter-item reliability analysis was performed to assess whether separate items, measuring similar entities, produced consistent results. Reliability was interpreted according to general agreement and also non-random agreement within or between questionnaire items. The reliability coefficient (R) was used to summarize non-random agreement above that expected by chance alone (complete concordance, R=1; chance concordance only, R=0). For items including qualitative (categorical) data, R was obtained using the 
statistic, and for quantitative (numerical scale) items the intraclass correlation method was used. Association between items was also investigated, using the Spearman correlation coefficient (rS). Thirdly, we compared response rates and prevalence estimates with results from other studies.
Part Three: general practice record review
In order to determine the validity of self-reported data in the questionnaire, individuals’ recall of knee pain and consultation for knee pain in the last 12 months was compared with data from general practice records. A total of 69% (136/197) of the available survey responders gave permission for their GP record to be reviewed. Three groups were drawn from this subsample: (group 1) all those who stated in the questionnaire both that they had knee pain and that they had consulted their GP for knee pain within the last 12 months (n=18); (group 2) a random sample of 20 people who stated in the questionnaire both that they had knee pain and that they had not consulted for knee pain in the last 12 months (10 in each practice); (group 3) a random sample of 20 people who stated in the questionnaire both that they did not have knee pain and that they had not consulted their GP for knee pain in the last 12 months (10 in each practice).
A total of 58 records were reviewed. An assessor (CJ) who was blinded to the above categories reviewed all records. Over half of the records (n=31) were validated by a second blinded assessor (KD) for the purposes of inter-rater agreement. Any consultation for knee pain, back or leg pain, or OA generally, was recorded on a record proforma. The exact date of each consultation was recorded.
To account for potential subject errors in timing recalled events to a 12-month period, three comparisons were made. First, questionnaire responses from all three sample groups above were compared with GP medical records for evidence of a consultation for knee pain within the same 12-month interval of the study (March 1998 to March 1999). Secondly, questionnaire responses from groups 1 and 2 above were compared with GP medical records for evidence of consultation for knee pain in the last 36 months (March 1996 to March 1999). Lastly, questionnaire responses for groups 1 and 2 above were compared with GP medical records for evidence of GP consultation for leg or knee pain in the last 36 months (March 1996 to March 1999).
The scores summarized agreement between self-reported and medical record data. The sensitivity and specificity of the KNEST questions against GP records were also calculated.
Randomization throughout the study was computer-generated, using SPSS version 9.0 (SPSS Inc, IL, USA). Statistical analysis was carried out using SPSS version 9.0 and STATA version 6.0 (STATA Corporation, TX, USA).
|
Results |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionReferences |
|---|
Questionnaire response and completion rates
A total of 198 people replied to the baseline questionnaire (adjusted response rate 85%). Completion rates for items in the KNEST are shown in Table 2
. The new knee pain question achieved 94% completion. All items on the SF-36, HADS and WOMAC were well completed, with ranges as follows: SF-36, 87.4–98%; HADS, 98–99%; WOMAC, 88–100%.
|
Test–retest reliability
Of the 80 responders who were mailed a retest questionnaire, a total of 59 (74%) returned the retest questionnaire. The observed agreement between the new knee pain question on baseline and repeat questionnaires was 90.9% (=0.82). Test–retest reliability for other items on the KNEST was also good (Table 3).
|
Test–retest analysis was also carried out for scores from the SF-36, HADS and WOMAC. Mean differences in each of the scales were close to zero, and good repeatability is further supported by significant intraclass correlations, most of which were above 0.6.
Inter-item reliability
Shading of pain on the manikin in any one of the four knee areas was compared with responses to the new knee pain question. Of the 82 questionnaires that had shading on the manikin, there were four with no answer to the knee pain question. Of the remaining 78 responses, 74 (95%) also replied ‘yes’ to the knee pain question.
Unilateral (right or left) or bilateral knee pain according to shading on the manikin was compared with the ‘knees affected’ question on the KNEST. There was a clear association between these items (rS=0.73, P<0.01).
Prevalence
A total of 187 respondents to the baseline questionnaire answered the new knee pain question. Of these responders, 84 people reported having knee pain, a 12-month period prevalence of 44.9% [95% confidence interval (CI)=37.8–52.0%]. There were no substantial age or gender differences. Most of these (67%) had had pain for more than 3 months during the previous 12 months.
Results of the GP record review
Inter-rater agreement between CJ and KD was 100%. Among group 3 of the record review subsample (i.e. those with no knee pain according to the questionnaire), there were no consultations for knee pain during the 12 months. Table 4 summarizes the findings for the other two record review subgroups. There was 90% sensitivity for the consultation question correctly identifying all consultations recorded, irrespective of whether they were within 12 or 36 months, about the knee or about some other lower limb pain. The specificity of the consultation question was 68% for knee pain consultations in the past year but higher if consultations during the previous 36 months or for the lower limb were included.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionReferences |
|---|
The aim of this pilot study was to develop and test a new knee pain screening tool to evaluate knee pain in primary care or for population needs assessment. We identified the need for a brief set of questions which would stand alone as a summary of knee pain severity and persistence and add previously unavailable information on intermittent pain, GP consultation and wider health-care use.
The full survey questionnaire used in this study (including the KNEST, SF-36, HADS and WOMAC) achieved a good response in the study population of people aged over 55 yr (85% at baseline), and completion rates were also high for all items on the questionnaire.
Items on the KNEST have content validity. Patients found the questions easy to understand and to complete. The questions were perceived to be appropriate and relevant to their knee pain. This was also true of the full composite questionnaire, except for the original WOMAC instruction page. The KNEST also displayed good construct validity, including correlation between related variables (e.g. test–retest and other internal consistency measures) and measures of external consistency (e.g. agreement between different methods of reporting).
Internal consistency
There was a high level of agreement (91%) between baseline and retest questionnaires with respect to the new knee pain screening question. Similarly, good test–retest agreement was found for manikin shading for knee pain. Other items in the KNEST and the composite tool also achieved good repeatability.
Prevalence figures for the new knee pain question were only 3% higher than those from the manikin, and there was also a significant association between the manikin and the KNEST in relation to the laterality of knee pain (unilateral/bilateral). However, the manikin-recorded 1-month period prevalence and the knee pain question refers to a 12-month period. Previous population surveys of hip, knee and low back pain have indicated that recall of a 12-month period of pain experience gives a similar figure to recall within a shorter period [22–24].
External consistency
Construct validity was examined through external consistency across different methods of measurement, including comparison with knee pain prevalence rates reported in other studies and comparison with consultations for knee pain extracted from GP medical records.
Prevalence of chronic knee pain
The 12-month period prevalence of knee pain in our study was 44.9% higher than reported in previous studies in the UK [6, 7]. However, these studies used definitions related to chronic pain. In our study, the prevalence of knee pain lasting for more than 3 months was similar to the figure from the other UK studies [6, 7]. The KNEST question about knee pain identifies chronic knee pain, but also intermittent and acute pain. If the purpose is to identify all knee pain sufferers in a population, the KNEST knee pain question may be a more suitable measure of identifying such cases in primary care or community settings, particularly as it can separate acute from chronic pain.
Agreement between self-report and GP records
Many studies have assessed the level of agreement between questionnaire data and medical records [25]. However, few studies have done this in relation to the knee, and there are no studies about recall of GP consultation for knee pain. Haapanen et al. [26] found poorer agreement between self-report and documented evidence in health centre and hospital medical records for hip and knee arthrosis (=0.48) than for other conditions, for example diabetes (=0.70) and cardiovascular diseases (=0.73–0.80). Reynoso and Cronan [27] found the correlation between self-report of HMO service utilization and actual service utilization amongst elderly people with OA to be low (0.366). Another study, by Kehoe et al. [28], reported that patient and physician recall of arthritis differed by 15%.
We have compared self-reports of knee pain and consultation for knee pain in the last 12 months with evidence of consultation extracted from GP medical records. Overall, agreement was good, but better if people's recall was compared with 36 months of the date of the record and if lower limb pain consultations were included. Such apparent over-reporting of GP consultation may be the result of simple inaccuracy. Symptoms may fluctuate over time, with periods of severe pain and disability followed by milder periods, and the frequency and duration of such events have been shown to contribute to recall [28]. However, it could also reflect the significance of past experiences of knee pain to the respondent: ‘Have you seen the doctor about trouble in that area in the recent past?’ seems to be a reasonable translation of ‘Have you consulted your GP about knee pain in the past 12 months?’
The second reason for recall error in this study concerns the location of pain. A small number of individuals reported consulting for knee pain, although reference to the knee could not specifically be found in the medical records. Instead, broader references to, for example, ‘leg pain’ were in evidence. This again may be error in patient recall or GP coding, but equally it may reflect different interpretations of pain location by GPs and their patients.
In this study, moderate reliability of recall about GP consultation in the last 12 months is reported (=0.47). By broadening the time period reviewed in the medical records only slightly (from a consultation within the last 12 months to a consultation within the last 15 months), reliability increased (=0.68). We conclude that, within reasonable bounds of interpretation, patient recall of GP attendance is a useful measure of GP attendance.
Conclusion
The Keele KNEST appears to be a valid tool for investigating the prevalence, severity and duration of knee pain and the use of health-care related to knee pain in the community and primary care. It performs well in relation to reliability and validity testing, and is a short screening instrument that is able to detect both acute and chronic knee pain in a population. It may be useful for Primary Care Groups and health authorities interested in knee pain as it can be used on its own or in combination with different validated questionnaires for other purposes. We are now conducting a population-based investigation of knee pain in relation to primary care provision using the full combination of instruments.
|
|
Acknowledgments |
|---|
The authors acknowledge the West Midlands New Blood Research Fellowship Committee, who awarded funding for this study, and the Haywood Foundation, who kindly supported the survey through a project grant. We thank Dr Krysia Dziedzic for helping with the general practice record review, and the doctors, staff and patients of the two health centres concerned.
|
Notes |
|---|
Correspondence to: C. Jinks, Primary Care Sciences Research Centre, 60 The Covert, Keele University, Keele ST5 5BG, UK.
|
References |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionReferences |
|---|
- Wilson MG, Michet CJ, Ilstrup DM, Melton LJ. Idiopathic symptomatic osteoarthritis of the hip and knee: a population-based incidence study. Mayo Clin Proc1990;65:1214–21.[Web of Science][Medline]
- Creamer P, Hochberg MC. Osteoarthritis. Lancet1997;350:503–9.[Web of Science][Medline]
- Felson DT, Zhang Y. An update on the epidemiology of knee and hip osteoarthritis with a view to prevention. Arthritis Rheum1998;41:1343–55.[Web of Science][Medline]
-
Dieppe P, Basler HD, Chard J et al. Knee replacement surgery for osteoarthritis: Effectiveness, practice variations, indicators and possible determinants of utilization. Rheumatology1999;38:73–83.
[Free Full Text] - Royal College of General Practitioners. Morbidity statistics from general practice. 1991–92. London: HMSO, 1995.
-
O'Reilly SC, Muir KR, Doherty M. Knee pain and disability in the Nottingham community: association with poor health status and psychological distress. Br J Rheumatol1998;37:870–3.
[Abstract/Free Full Text] -
McAlindon TE, Cooper C, Kirwan JR, Dieppe PA. Knee pain and disability in the community. Br J Rheumatol1992;31:189–92.
[Abstract/Free Full Text] - von Korff M, Ormel J, Keefe F, Dworkins S. Grading severity of chronic pain. Pain1992;50:133–49.
- Hilderbrandt J, Franz CE, Choroba-Mehnen B, Temme M. The use of pain drawings in screening for psychological involvement in complaints of low back pain. Spine1988;13:681–5.[Medline]
- Margolis RB, Chibnall JT, Tait RC. Test–retest reliability in the pain drawing instrument. Pain1988;33:49–51.[Web of Science][Medline]
- Love A, Leboeuf C, Crisp TC. Chiropractic chronic low back pain sufferers and self report assessment methods. Part 1. A reliability study of the visual analogue scale, the pain drawing and the McGill questionnaire. J Manipulative Physiol Ther1989;12:21–5.[Web of Science][Medline]
- Brazier JE, Harper R, Jones NMB et al. Validating the SF-36 health survey questionnaire: new outcome measure for primary care. Br Med J1992;305:160–4.
- Garratt AM, Ruta DA, Abdalla MI, Buckingham JK, Russell IT. The SF-36 health survey questionnaire: an outcome measure suitable for routine use within the NHS? Br Med J1993;306:1440–4.
-
Brazier J, Harper R, Munro J, Walters J, Snaith M. Generic and condition-specific outcome measures for people with osteoarthritis of the knee. Rheumatology1999;38:870–7.
[Abstract/Free Full Text] - Herrmann C. International experiences with the Hospital Anxiety and Depression Scale—A review of validation data and clinical results. J Psychosom Res1997;42:17–41.[Web of Science][Medline]
- Bellamy N. WOMAC Osteoarthritis Index: A users guide. London, Ontario: University of Western Ontario, 1996.
-
Hayes V, Morris J, Wolfe C, Morgan M. The SF-36 health survey questionnaire: is it suitable for use with older adults. Age Ageing1995;24:120–5.
[Abstract/Free Full Text] - Brazier JE, Walters SJ, Nicholl JP, Kohler B. Using the SF-36 and Euroqol on an elderly population? Qual Life Res1996;5:195–204.[Web of Science][Medline]
- Hobson JP, Meara RJ. Is the SF-36 health survey questionnaire suitable as a self-report measure of the health status of older adults with Parkinson's disease? Qual Life Res1997;6:213–6.[Medline]
-
Lyons RA, Perry HM, Littlepage BN. Evidence for the validity of the Short-Form 36 Questionnaire (SF-36) in an elderly population. Age Ageing1994;23:182–4.
[Abstract/Free Full Text] - Altman D. Practical statistics for medical research. London: Chapman & Hall, 1991:404.
- Croft P, Papageorgiou A, McNally, R. Low back pain: health care needs assessment. Second series. Oxford: Radcliffe Medical Press, 1997:12.
-
Walsh K, Gruddas M, Coggon D. Low back pain in eight areas of Britain. J Epidemiol Community Health1992;46:227–30.
[Abstract/Free Full Text] - Mason V. The prevalence of back pain in Great Britain. Office of Population Censuses and Surveys, Social Survey Division. London: HMSO, 1994.
-
Harlow S, Linet M. Agreement between questionnaire data and medical records. Evidence of the accuracy of recall. Am J Epidemiol1989;129:233–48.
[Free Full Text] - Haapanen N, Miilunpalo S, Pasanen M, Oja P, Vuori I. Agreement between questionnaire data and medical records of chronic diseases in middle-aged and elderly Finnish men and women. Am J Epidemiol1997;145:762–9.
- Reynoso L, Cronan T. Are self reports of health care utilization valid? Ann Behav Med1995; 17(Suppl.):S197.
-
Kehoe R, Wu SY, Lesske M, Chylack L. Comparing self reported and physician reported medical history. Am J Epidemiol1994;139:813–8.
[Abstract/Free Full Text]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  C. Jinks, K. P. Jordan, M. Blagojevic, and P. Croft Predictors of onset and progression of knee pain in adults living in the community. A prospective study Rheumatology, March 1, 2008; 47(3): 368 - 374. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Hill, K. Dziedzic, E. Thomas, S. R. Baker, and P. Croft The illness perceptions associated with health and behavioural outcomes in people with musculoskeletal hand problems: findings from the North Staffordshire Osteoarthritis Project (NorStOP) Rheumatology, June 1, 2007; 46(6): 944 - 951. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Dziedzic, E. Thomas, S. Hill, R. Wilkie, G. Peat, and P. R. Croft The impact of musculoskeletal hand problems in older adults: findings from the North Staffordshire Osteoarthritis Project (NorStOP) Rheumatology, June 1, 2007; 46(6): 963 - 967. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Jinks, K. Jordan, and P. Croft Osteoarthritis as a public health problem: the impact of developing knee pain on physical function in adults living in the community: (KNEST 3) Rheumatology, May 1, 2007; 46(5): 877 - 881. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R Duncan, G Peat, E Thomas, E Hay, I McCall, and P Croft Symptoms and radiographic osteoarthritis: not as discordant as they are made out to be? Ann Rheum Dis, January 1, 2007; 66(1): 86 - 91. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. C. Duncan, E. M. Hay, J. Saklatvala, and P. R. Croft Prevalence of radiographic osteoarthritis--it all depends on your point of view Rheumatology, June 1, 2006; 45(6): 757 - 760. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. R. Underwood Community management of knee pain in older people: is knee pain the new back pain? Rheumatology, January 1, 2004; 43(1): 2 - 3. [Full Text] [PDF]
|
|
|
|
|
|
C. Jinks, K. Jordan, B. N. Ong, and P. Croft A brief screening tool for knee pain in primary care (KNEST). 2. Results from a survey in the general population aged 50 and over Rheumatology, January 1, 2004; 43(1): 55 - 61. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||