In the UK, primary care is commonly the point of entry into the health care system for people with a new symptom or illness and the major source of continuing care for chronic conditions. Musculoskeletal problems are one of the most common reasons for seeking primary care, with estimates of up to 20% of adults consulting their general practitioner with a musculoskeletal problem over the course of a year[1, 2].

In order to plan primary health care and monitor the clinical course of problems presented to primary care, increasing use is made of routine electronic recording of the reasons people consult. There is much interest in estimating the level of health care need for musculoskeletal disorders, and then in identifying suitable and effective interventions. Patients with musculoskeletal problems often present to primary care with a regional symptom, such as back, knee or shoulder pain. Many musculoskeletal symptoms cannot be labeled with a diagnosis initially and GPs often find it more useful to work with a regional pain label than a complex diagnostic label. Region specific management is common in primary care[3]. Randomised controlled trials in primary care of management of regional pain symptoms, as opposed to diagnoses, are increasing (e.g. [4–6]), leading to requirements to estimate the number of people who consult who meet the criteria for recruitment. Epidemiological studies are increasingly using routinely recorded primary care data to model occurrence, management and outcome of consultation for regional and widespread musculoskeletal problems, and so also need to identify patients with specific regional problems[7–10].

Generally, the increasing use and quality of electronic records in primary care means that it should be easier to determine the consultation prevalence related to specific types of morbidity. UK databases such as the Fourth National Survey of Morbidity in General Practice (MSGP4) conducted by the Royal College of General Practitioners (RCGP) last carried out in 1991/2 [11] and the linked annual RCGP Weekly Returns Service (WRS) reports [2] provide figures on annual prevalence of consultation to primary care. However, in terms of musculoskeletal disorders, these problems are often either given diagnostic labels such as osteoarthritis or rheumatoid arthritis, or labelled as non-specific disorders. Systems of morbidity coding do not lend themselves easily to the identification of prevalence of consultation for individual regional presenting symptoms, because the principles of coding in primary care do not neatly or necessarily reflect a regional classification of symptoms. The Read Code hierarchy, a common system for recording morbidity in UK primary care, is structured into diagnostic chapters (for example, Musculoskeletal diseases, Mental disorders, Circulatory disorders)[12]. The most common musculoskeletal problems diagnosed in primary care by the RCGP WRS in 2007 within the Musculoskeletal Read Code Chapter were listed as "Back disorders unspecified other", "Pain in Limb", "Joint Disorder Other Unspecified", "Osteoarthritis & Allied Disorders" and "Peripheral Enthesopathies & Allied Syndromes"[2].

Most musculoskeletal morbidity codes fall within the Musculoskeletal Disorders Chapter of coding systems but may also fall into the symptoms Chapters or the injury Chapter. The Read Code hierarchy has, for example, over a hundred knee problem codes under different diagnostic chapters[13]. Widespread or generalised problems are often labelled as multiple individual problems in primary care[10], and labels such as fibromyalgia are rarely used[14]. Previous studies of regional problems in primary care which have used general practice consultation databases have been limited by either concentrating on a subset of codes (generally from within the Musculoskeletal Disorders Chapter), had small samples, have scrutinised written records, or restricted analysis to specific sites or ages[9, 15–19]. They have generally not allowed comparison of the relative prevalence of primary care consultation across multiple body regions and therefore the extent to which presentation of musculoskeletal problems by region varies is unknown.

The aims of this study were to determine the annual consultation prevalence for musculoskeletal problems in primary care, classified by body region, and to develop a resource for primary care researchers working in this field.

The study was performed by first identifying all musculoskeletal Read Codes and classifying them into body regions, and then applying the classification to a primary care consultation database.

This study has highlighted the frequency of musculoskeletal problems in primary care. One in seven of all recorded consultations in the twelve study practices during 2006 was for a musculoskeletal problem. One in four of the registered population consulted for a musculoskeletal problem in that year, rising to more than one in three of older adults. The back was the most common reason for consultation, followed by the knee, chest and neck; except in children where the foot was the predominant regional problem.

Overall consultation prevalence of musculoskeletal problems in adults for codes within the Musculoskeletal Disorders Read Code Chapter was similar to that found in a Swedish study[15]. The finding that females were more likely than males to consult primary care for regional musculoskeletal problems is consistent with previous studies examining consultation prevalence for musculoskeletal symptoms[16], chronic pain[15], lower limb problems[17], shoulder problems[9], and neck and upper limb problems[18]. The one region tending towards male susceptibility (although not statistically significant) was the elbow, which also rose in prevalence in working age groups and may be related to the association of elbow problems with manual occupations.

Other studies have concentrated on specific regions of the body, and have not shown the relative prevalence rates across regions. Consultation prevalence for lower limb problems (10%) in our study was higher than that reported by van der Waal (6%)[17], probably reflecting our inclusion of codes outside of the Read musculoskeletal Chapter. Consultation prevalence for shoulder problems in adults was similar to that (2.4%) reported for adults by Linsell et al who also reported higher prevalence in females and older age groups[9]. Consultations for the shoulder as a proportion of all consultations were also similar to our study (0.71%).

Widespread problems were common with 5% of the registered population consulting for several regions, rising to 10% of those aged 45 and over. Consulting for other regions was seen in a Finnish study in those consulting initially with neck or shoulder problems[22]. There is also evidence from population surveys that multiple site problems dominate in persons with chronic musculoskeletal pain with only a quarter of those reporting chronic pain having it in a single site[23]. That survey also found that females and the older aged were more likely to report multiple pain sites and this is consistent with the consultation patterns we have identified. Although our study was restricted to a period of one year, it suggests a substantial amount of chronic and widespread pain in primary care, a problem known to relate to poorer general health[10]. However, in younger age groups particularly, most encounters are coded as single site problems, and it is only in the older age groups that more generalised codes are recorded by the health care professional.

About one in four musculoskeletal consultations were given codes outside the musculoskeletal (Chapter N) chapter. These included symptoms such as knee pain and injuries such as fractures. Some GPs may prefer to use a symptom code until they can clearly make a diagnosis. Using only codes within the titled musculoskeletal chapter, would lead to an underestimate of the musculoskeletal problems seen in primary care.

This study was performed within a high quality dataset which included practices who undergo annual assessments of their morbidity coding, have reached levels of 97% of contacts being given a code, and give comparable general consultation rates to national databases[1, 20]. The grouping of codes to regions was performed by independent clinical practitioners, prior to extraction of consultations from CiPCA. This approach to using morbidity codes to map musculoskeletal regions offers a standardised method to investigate the occurrence, management and outcome of regional problems in other primary care databases using Read codes. Further, alongside our consultation prevalence figures they offer a basis for researchers planning trials and epidemiological studies in primary care.

Our study data was drawn from a registered population of over 100,000 persons but only a small number of practices in one area of UK which is slightly more deprived than England as a whole. However, the CiPCA database has been shown to give comparable age-gender standardised consultation prevalence figures for musculoskeletal problems overall, and for osteoarthritis, rheumatoid arthritis and arthralgia as the larger national RCGP WRS database[1]. Whilst the classification into regions is most relevant to the UK where Read Codes are used, the consultation prevalence figures are likely to be relevant to countries with similar primary care systems.

About 6% of the identified musculoskeletal consultations could not be attributed to a region or to being a widespread problem as the recorded codes lacked enough detail. There is no reason to suggest GPs were less likely to label one region than another but our figures may be a slight underestimate of the true consultation prevalence for individual regional problems. A further issue is the overlapping of regions. For example, overlap has been reported previously in recording of neck and shoulder complaints[19]. Knee pain may also be included in a consultation recorded as lower leg, leading to an underestimate of the consultation prevalence for the knee. However the identified consultation prevalence for lower leg was very low.

It is possible that some of the symptom codes outside of the defined Musculoskeletal Chapter may not have the diagnostic features of musculoskeletal problems. The chest was a common reason for consultation and consultations for chest problems may not always be regarded as a musculoskeletal problem[24]. The predominant chest codes identified in our study was a symptom code labelled "Chest pain" and its daughter code "Musculoskeletal chest pain". It is possible the label of "Chest pain", as well as including musculoskeletal problems, may include problems that are not musculoskeletal. However, it would be expected that if the symptom was cardiac or gastrointestinal in nature, for example, then it would have been coded under these Chapters and hence would not have been identified by our classification, unless the symptom code is being used until a firm diagnosis is made. There is evidence that the main underlying reason for all chest pain seen in primary care is musculoskeletal with separate studies suggesting that musculoskeletal reasons account for between 20% and 49% of chest pain cases[24–26]. This would all suggest the majority of chest pain consultations in our study were musculoskeletal in nature.

We based annual consultation for widespread problems on two different definitions: counts of body sectors from separate consultations and single consultations labelled with a generalised code. Definitions of widespread pain vary and include use of widespread morbidity codes, specific diagnoses such as fibromyalgia, and definitions based on varying numbers and areas of the body. Some criteria are more restrictive than the one we have used[27, 28]. This makes comparisons between studies difficult. Our definition indicates people with pain in at least two distinct regions.

As with all studies set within primary care, this study can only measure problems for which health care is sought. The general population prevalence of problems will be higher as many people will have regional pain but do not consult.

The use of a classification system based on body regions rather than diagnosis highlights the extensive and varied musculoskeletal workload in primary care, and the extent of multiple regional problems. Musculoskeletal problems need to be recognised as a major constituent of general practice, particularly the extent of problems in multiple body regions, highlighted by this new classification system. With the increased use of electronic recording in recent years, this new classification of regional problems provides a potential resource for researchers to standardise randomised controlled trials of regional pain in primary care and identify eligible participants from morbidity records, or to assess the primary care management and outcome of regional musculoskeletal problems.