MSK Calculator FAQ

What is the Musculoskeletal Calculator?

Versus Arthritis has partnered with Imperial College London to develop the Musculoskeletal Calculator, a tool designed to produce prevalence estimates (i.e. how many people have a condition) for musculoskeletal (MSK) conditions.

What estimates are available?

At this stage estimates are available for osteoarthritis of the hip and knee (general (i.e. total)/severe), back pain (general (i.e. total)/severe) and rheumatoid arthritis for England and Scotland. Estimates for Wales and high risk of fragility fractures for all countries are expected to be available soon. Prevalence estimates for Northern Ireland are not possible due to a lack of local risk factor data.

Data is available at local authority (LA), Clinical Commissioning Group (CCG), or Health Board (HB) level.

Location Osteoarthritis (hip/knee) Back pain Rheumatoid arthritis High risk of fragility fracture
Scotland YES - LA, HB YES - LA, HB YES - HB NO - TBC
Northern Ireland NO - N/A NO - N/A NO - N/A NO - N/A

Why have we developed the Musculoskeletal Calculator?

Before the Musculoskeletal Calculator was developed, there were no local estimates available on the number of people with these common MSK conditions. By not measuring these conditions, it became difficult for local planners to take them into account when planning their services and risked the conditions being overlooked. A lot of power in the healthcare system, in particular public health, rests at a local level, so allowing estimates to be produced at this level was especially important. Local authorities can have an impact on MSK risk factors like physical activity, and general practices can improve diagnosis by finding and accurately recording new cases of disease.

Why are they estimates?

The data provided are estimates because it is not presently possible to access the health data records required to determine how many people have an MSK condition. Most of the treatment for MSK conditions occurs in primary care and this data is not consistently collected by the NHS. The numbers and percentages derived by the Musculoskeletal Calculator are based on a statistical model. Versus Arthritis is campaigning for more data on musculoskeletal conditions to be made available.

What data sources were used to calculate local estimates?


Condition Source of numerator Description of source Source of denominator
Osteoarthritis (hip/knee) The English Longitudinal Study of Ageing (ELSA) Wave 1 (2000/01) – Wave 5 (2010/11) The ELSA is a large multicentre and multidisciplinary study of people aged 50 and over and their younger partners, living in private households in England.

This survey was chosen because the sample used in ELSA was designed to be nationally representative and osteoarthritis is most common in those aged over 45 years of age.

The survey uses patient-reported doctor diagnosed disease criteria; it includes questions concerning limitations with activities of daily living allowing us to differentiate disease severity.
ONS mid-year population estimates for 2012.

People aged 45 and over only.
Back pain Health Survey for England (HSE) 2011 The HSE is an annual survey of adults and children, representative of the whole population at both national and regional level designed to measure health and health related behaviours.

This survey was chosen because it contains several questions around pain or discomfort and specifically chronic pain, which allows us to differentiate disease severity.
ONS mid-year population estimates for 2012.

People of all ages.
Rheumatoid arthritis Clinical Practice Research Datalink (CPRD)-extracted 23/01/15 The CPRD is governmental, not-for-profit research service that provides anonymised primary care records for public health research.

This database was chosen over a survey, due to the breadth of data available to identify possible cases, including clinical, prescribing and test data.>
ONS mid-year population estimates for 2015.

People aged 16 and over.
High risk of fragility fracture TBC TBC TBC


How are the estimates calculated?

The numbers and percentages from the Musculoskeletal Calculator are estimates based on statistical models. A lay summary to the methods used to develop these models are as follows:

  • Identify risk factors: A literature review was conducted via the MedLine database and supplemental references provided by Versus Arthritis to identify risk factors and comorbidities (if applicable) associated with an outcome (i.e. osteoarthritis of the hip/knee).
  • Choose data source: Several existing data sources were analysed (i.e. HSE, ELSA, GLS, CPRD) to identify which dataset best records the outcome in question (i.e. osteoarthritis of the hip/knee). The final data source was selected based on its data qualities (i.e. validity, completeness, timeliness etc.). For example, ELSA is more specific than HSE or GLS, because it allows a specific type of arthritis to be recorded, which was an important reason for selecting it when examining outcomes (i.e. osteoarthritis of the hip/knee)
  • Data mapping: Risk factors and comorbidities (if applicable) identified in the literature review were mapped out in the dataset to determine 1.) the proportion of the missing data 2.) need for multiple imputation to make up for missing data and 3.) inclusion of risk factor data in regression model.
  • Regression modelling and validation: Baseline characteristics and regression analysis was conducted via Stata 11 to identify the relationship between risk factors and comorbidities (if applicable) available in the dataset and the outcome variable (i.e. osteoarthritis of the hip/knee). Sensitivity and specificity were calculated using the area under the ROC curve to validate results.
  • Synthetic estimation: application of the model to small population data: Derived Odds Ratios from regression modelling were then used to estimate prevalence of the condition in small population subgroups (i.e. General Practice and Middle Layer Super Output Area (MLSOA) populations) using local population breakdowns for each risk factor where available.

How accurate are the estimates?

The numbers and percentages from the Musculoskeletal Calculator are estimates based on statistical models. However, the data sources used to produce these models and estimates are designed to give a representative picture for the whole of England. National bodies routinely use survey data: for example, ELSA is jointly funded by UK government departments and the National Institute on Aging (USA) and “helps the government plan health care services and pensions systems to accurately meet the needs of older people”.

How did we identify data on the risk factors?

A scientific literature review was conducted to identify key risk factors which increase/decrease the chances of developing the condition in question. For example, there’s a clear link between obesity and osteoarthritis of the knee. Answers to the questions in the survey’s relating to these risk factors were then analysed. For example, someone’s response to questions around weight (BMI) were used to examine the relationship between obesity and osteoarthritis of the knee.

As with all datasets, some data was missing, in which specific statistical techniques were used that take into account missing data.

What locally available risk factor data was used to calculate the local prevalence estimates?

Osteoarthritis (hip/knee) Back pain Rheumatoid arthritis
Age group 
Socioeconomic factors 
Smoking status 
Leisure physical activity 
Membership at gym/sports club 
Age group 
Socioeconomic factors 
Smoking status 
Deprivation (IMD) 

How can you know how many people have osteoarthritis or back pain if so many go undiagnosed?

There is no simple test for conditions like osteoarthritis or back pain, diagnoses vary greatly between individual GPs and across general practices. GPs often use symptom labels such as 'knee pain' rather than a diagnostic label of osteoarthritis and by the time someone has a GP appointment, their back pain may have subsided.

Because of this the Musculoskeletal Calculator defined two main groups as having osteoarthritis of the hip or knee. Survey respondents who reported that they have doctor-diagnosed osteoarthritis and/or respondents who reported consistent hip or knee pain.

Similarly, the Musculoskeletal Calculator defined someone as having back pain as survey respondents who said they are 1. currently troubled by pain or discomfort, 2. the pain or discomfort lasted for more than three months, and 3. the site of the pain was identified as “back pain”.

What's the difference between (Total) and (Severe) osteoarthritis and back pain?

Total includes all cases of osteoarthritis of the hip or knee or back pain. Severe includes cases that would require special attention, and additional resources, from healthcare providers and commissioners.

Severe osteoarthritis (hip/knee) and back pain are defined as follows:

Condition Definition
Severe osteoarthritis (hip/knee) This severity of osteoarthritis was determined using two survey questions:
  • 'Severity of pain most of the time'
  • 'Difficulty walking ¼ mile unaided'.
Respondents were deemed to have 'severe' osteoarthritis if their answers included any one of the following statements:

They have severe pain most of the time (as opposed to 'mild' or 'moderate').

They're unable to walk ¼ mile unaided (as opposed to 'no', 'some' or 'much difficulty').

They've previously undergone hip or knee replacement due to arthritis.
Back pain The severity of back pain was determined using the Chronic Pain Grade based on GCPS version 2.0:

Grade 0 - no pain problem 
Grade I – Low intensity, low interference 
Grade II – High intensity 
Grade III – Moderate interference (limiting) 
Grade IV – Severely interference (limiting) 

Respondents were deemed to have severe back pain if they scored a chronic pain grade of II, III, or IV.

Where can I learn more about what you are doing for people with a musculoskeletal condition?

Versus Arthritis is constantly seeking to make sure that the needs of people with musculoskeletal conditions are considered when decisions are made about the healthcare system. You can learn more about what we’re asking politicians to do by reading our manifesto.

How can I find out more?

If you would like to know more, you can access and download the technical documents below. If you have any other queries please contact