Trends, geographical variation and factors associated with prescribing of gluten-free foods in English primary care: a cross-sectional study.
Walker AJ, Curtis HJ, Bacon S et al. BMJ Open 2018;8:e021312. Doi:10.1136/bmjopen-2017-021312
Adherence to a strict gluten-free diet (GFD) is the only effective treatment for coeliac disease and prescribing of gluten-free products for these patients may be associated with better adherence. Currently, there is extensive discussion as to whether gluten-free prescribing is a cost-effective intervention with wide variation in prescribing policies between Clinical Commissioning Groups (CCGs).
Adherence to a strict gluten-free diet (GFD) is the only effective treatment for coeliac disease and prescribing of gluten-free products for these patients may be associated with better adherence. Currently, there is extensive discussion as to whether gluten-free prescribing is a cost-effective intervention with wide variation in prescribing policies between Clinical Commissioning Groups (CCGs).
The recent NICE Quality Standard stated “healthcare professionals should highlight if gluten-free food products are available on prescription to help people maintain a GFD” which demonstrates the risk of inequality of access to these products in England. In comparison, recommendations for the prescribing of gluten-free products in Northern Ireland, Scotland and Wales are broadly in line with the national prescribing guidelines that exist.
The aim of this retrospective cohort study was to describe trends over time and national variation in gluten-free prescribing as well as assessing factors associated with the rate of prescribing of gluten-free products in English general practices. Annual prescription cost analysis (PCA) data was used to describe long-term trends in gluten-free prescribing (between 1998-2016) as well as other data published by NHS Digital. Prescribing data was also matched with publicly available data from Public Health England allowing stratification of the analysis to look at reasons for variation in prescribing practices at practice level. All standard English practices, labelled as a GP practice within the data (n=8185), were included within the analysis, however, prescribing in non-standard settings, such as prisons was excluded. In addition, those practices no longer active, those without a 2015/16 Quality Outcomes Framework (QOF) score and those with a list size of <1000 were also excluded (n=558). The use of inclusive data such as this reduced the likelihood of obtaining a biased sample.
Whilst overall prescribing was found to have risen from 1998 to 2010, a substantial decline since this peak has been observed. A high level of variation in rate of prescribing between practices was seen, some of which is driven at CCG level, where there is also a great deal of variation observed. This variation remained fairly similar over time. The factors affecting the rate of gluten-free prescribing were examined through modelling. These were found to include the level of deprivation in an area, with a significantly lower rate of prescribing seen in the most deprived areas, and practices performing poorly on a previously developed composite measure of prescribing quality being more likely to prescribe gluten-free foods. Age distribution and practice size were also important in determining variation in the gluten-free prescribing rate amongst other factors.
Other possible measures for variation in prescribing that the study could not measure included incidence of coeliac disease. National variation in the regional prevalence of coeliac disease has been reported in previously published research, however, this is too broad a level to be meaningfully adjusted for in the analysis. In addition, the reasons for variation in prevalence would be similar to those for variation in gluten-free prescribing, for example, index of multiple deprivation is known to be associated with the incidence and prevalence of coeliac disease. CCGs were found to be a significant driver of variation, with large variation observed at CCG level, likely due to the differences in policies. This suggests GP practices are responsive to CCG prescribing guidance, at least in relation to gluten-free products.
In summary, the prescribing of gluten-free foods is rapidly declining which may be viewed as a positive change by some allowing resources to be more effectively used elsewhere. However, patient groups, such as Coeliac UK, are strong advocates for the continuation of gluten-free prescribing to help support adherence, with particular concern for vulnerable populations being able to source appropriate foods for their condition without prescriptions. Therefore, any potential further reductions in prescribing would be controversial. It is clear that the level of variation in this prescribing area is very high; this variation appears to exist largely without good reason and is determined, in the main, by decisions and preferences of clinicians and local health services.
Link to full paper
The aim of this retrospective cohort study was to describe trends over time and national variation in gluten-free prescribing as well as assessing factors associated with the rate of prescribing of gluten-free products in English general practices. Annual prescription cost analysis (PCA) data was used to describe long-term trends in gluten-free prescribing (between 1998-2016) as well as other data published by NHS Digital. Prescribing data was also matched with publicly available data from Public Health England allowing stratification of the analysis to look at reasons for variation in prescribing practices at practice level. All standard English practices, labelled as a GP practice within the data (n=8185), were included within the analysis, however, prescribing in non-standard settings, such as prisons was excluded. In addition, those practices no longer active, those without a 2015/16 Quality Outcomes Framework (QOF) score and those with a list size of <1000 were also excluded (n=558). The use of inclusive data such as this reduced the likelihood of obtaining a biased sample.
Whilst overall prescribing was found to have risen from 1998 to 2010, a substantial decline since this peak has been observed. A high level of variation in rate of prescribing between practices was seen, some of which is driven at CCG level, where there is also a great deal of variation observed. This variation remained fairly similar over time. The factors affecting the rate of gluten-free prescribing were examined through modelling. These were found to include the level of deprivation in an area, with a significantly lower rate of prescribing seen in the most deprived areas, and practices performing poorly on a previously developed composite measure of prescribing quality being more likely to prescribe gluten-free foods. Age distribution and practice size were also important in determining variation in the gluten-free prescribing rate amongst other factors.
Other possible measures for variation in prescribing that the study could not measure included incidence of coeliac disease. National variation in the regional prevalence of coeliac disease has been reported in previously published research, however, this is too broad a level to be meaningfully adjusted for in the analysis. In addition, the reasons for variation in prevalence would be similar to those for variation in gluten-free prescribing, for example, index of multiple deprivation is known to be associated with the incidence and prevalence of coeliac disease. CCGs were found to be a significant driver of variation, with large variation observed at CCG level, likely due to the differences in policies. This suggests GP practices are responsive to CCG prescribing guidance, at least in relation to gluten-free products.
In summary, the prescribing of gluten-free foods is rapidly declining which may be viewed as a positive change by some allowing resources to be more effectively used elsewhere. However, patient groups, such as Coeliac UK, are strong advocates for the continuation of gluten-free prescribing to help support adherence, with particular concern for vulnerable populations being able to source appropriate foods for their condition without prescriptions. Therefore, any potential further reductions in prescribing would be controversial. It is clear that the level of variation in this prescribing area is very high; this variation appears to exist largely without good reason and is determined, in the main, by decisions and preferences of clinicians and local health services.
Link to full paper
www.drschaer-institute.com