Prevalence Of MS In General Populations

Sara L Thomas,^a Rachael Williams,^b Tim Williams,^b Andrew J Hall.^a

^a Department of Epidemiology & Population Health, London School of Hygiene & Tropical Medicine
^b The General Practice Research Database

Background
Estimates from a decade ago suggested that 85,000 individuals were living with an MS diagnosis in the UK, but the accuracy of this figure is unclear. We analysed UK population‐based data to obtain updated estimates of the lifetime prevalence of MS and the number of individuals living with MS.

Methods
We analysed anonymised electronic medical records from the General Practice Research Database (GPRD), which covers approximately 6% of the UK population. We identified all patients registered with a GPRD
practice between 2000 and 2008 with a ‘definite MS’ diagnostic code anywhere in their medical record.

Prevalence of MS was calculated as the proportion of the total GPRD population who had ever had an MS diagnosis, stratified by age, sex, geographical region and calendar year. Numbers for 2007 were scaled up
to the 2007 UK population, using data from the Office for National Statistics. Estimates were adjusted for possible misdiagnosis/miscoding of MS in the GPRD, using 1) a positive predictive value (PPV) of 82% for MS
diagnostic codes, ascertained in a previous GPRD validation study, and 2) the proportion of MS patients who had evidence of neurologist referral and/or hospital feedback. Analyses were repeated among individuals registered with a practice for ≥1 year, to account for delayed reporting of historical MS diagnoses by practices.

Results
Crude annual prevalence of MS increased from 17.2/104 (95% CI:16.7‐17.7) in 2000 to 19.8/10⁴ (95% CI:19.3‐20.3) in 2008. In 2007, prevalence was 28.1/10⁴ (95% CI:27.3‐28.9) among females and 10.8/10⁴
(95% CI:10.3‐11.3) among males, and was highest among those aged 55‐64 years. Marked geographical variation was seen, with highest prevalence in 2007 in the Scottish Highlands/Islands (49.7/10⁴, 95% CI:31.8‐73.9) and lowest in Greater Glasgow/Clyde (12.8/10⁴, 95% CI:8.8‐18.1). After applying the percentage of male (72%) and female (75%) patients with neurologist referral/hospital feedback, the estimated number of individuals living with an MS diagnosis in 2007 was 88,760 (range:85,820‐91,770).
Using a PPV of 82% for an MS diagnostic code, the estimated number of individuals living with MS was 98,110 (range:94,820‐101,450). Estimates were slightly higher when the 2007 GPRD population was restricted to those with ≥1 year’s follow‐up.

Conclusion
This analysis provides updated estimates for the number of people living with MS in the UK. The apparent increase in MS prevalence over time is likely to be due in part to increased ascertainment of MS cases and improved recording of MS diagnoses. Over‐estimation of the number of MS cases may have occurred, due to miscoding of tentative MS diagnoses as ‘definite MS’ cases. Conversely, MS cases may have been underestimated,
due to continuing under-ascertainment and under-recording of MS cases in general practice data, and to possible inaccuracies in the estimates of the PPV of an MS diagnostic code and the proportion of individuals with neurology referrals/hospital feedback. Future estimates of lifetime MS prevalence will be facilitated by ongoing improvements in general practice recording of diagnoses and by increasing data linkage between general practice and other medical records.

Commenting on the study, Helen Yates, Multiple Sclerosis Resource Centre Chief Executive said, “ MSRC has been using a figure of 100,000 for the past 12 months.  We have always believed the number of people with MS was higher than the very old 85,000 figure.  It is good to see some further indication that this figure is much more likely to be accurate”

Source: MSRC,  London School of Hygiene & Tropical Medicine, MSS (12/06/09)