Has the nGMS contract improved diabetes care in general practice?

In 2004 the new GMS contract was negotiated. Under this contract GPs are offered financial rewards for achieving clearly defined targets for the management of chronic conditions. The number of points on offer for diabetes reflects the growing prevalence and burden of this condition (DoH, 2003b). 

This prospective longitudinal observational study collected data in 2001, 2003 and 2007 from nine practices in the Guildford and Waverly area in the south of England. Figure 1 shows the timing of data collection in relation to national initiatives. Data were collected for all patients registered at each practice. The records studied were 91499 in 2001, 92996 in 2003 and 97511 in 2007. Informed consent was obtained from practices prior to data collection. Anonymised data were extracted using MIQUEST queries to provide information on predefined indicators of diabetes care: 

• absolute value for HbA_1c 
• cholesterol 
• blood pressure 
• smoking status 
• BMI.

To be included, data must have been recorded in the previous 15 months, in line with nGMS targets (DoH, 2003b). Demographic data were also collected. People with diabetes were identified by the presence of a diagnostic computer read code, and the time since diagnosis was calculated as the time since the first diagnostic diabetes code was entered. No restriction was placed on age range.

Under the nGMS contract, practices are able to exclude patients from certain indicators – if, for example, a patient refuses monitoring. Exception reporting was not taken into account as this information was not available for the earlier years of data collection, and because there is evidence that practices with high exception reporting rates perform better (Doran et al, 2006).

Statistical analyses were performed using SPSS 14.0 software. Data were compared using independent sample t-tests, or Mann–Whitney U tests for skewed data. Chi-squared tests were used for categorical data.

Results
Between 2001 and 2007 the recorded prevalence of diabetes rose from 1.89% to 2.55%. Table 1 shows demographic data for the population with diabetes. Table 2 shows mean values and the percentage of people with diabetes who have had each parameter recorded in a given time period, together with the level of statistical significance.

There was a significant increase in the recording of all parameters over the 15 months preceding data collection, although the recording of blood pressure dropped between 2001 and 2003 before rising from 2003 to 2007. A substantial decrease in HbA_1c occurred, from 8.1% in 2001 to 7.2% in 2007. During this period the average blood pressure fell from 143.4/79.3mmHg to 135.7/76.4mmHg. Mean cholesterol fell from 5.0mmol/l to 4.5mmol/l over the course of the study. The average BMI of the population with diabetes rose slightly but significantly from 28.6kg/m² to 29.4kg/m². In addition, there was a marked decrease in the number of people recorded as being current smokers, from 25.1% in 2001 to 13.5% in 2007.

Discussion
This prospective longitudinal observational study shows that standards of care and data recording have improved substantially since 2001. By 2003 GPs were well placed to respond to the nGMS contract, and standards have continued to rise with its implementation. In many cases this has led to a significant improvement in modifiable risk factors for the long–term complications of diabetes. Our results add to the evidence base suggesting that financial incentives may be an effective way of improving quality of care (Chaix-Couturier et al, 2000). The increase in prevalence of diabetes demonstrated by this study is in line with that seen in larger local and national epidemiological studies (Amos et al, 1997; Gatling et al, 1998). It is likely that to some degree the increase in prevalence is due to better detection and improved data governance. However, this in itself is still significant, as it means that more people are being diagnosed and offered appropriate management.

This decade has been a time of rapid change in health care in the UK, making it impossible to attribute improvements to any one initiative. Standards of care certainly have improved with the introduction of the nGMS contract, however, standards were already improving prior to its introduction. As the new contract did not have a geographically staggered introduction it is difficult to confidently isolate and assess its impact. It may be that the improvements in quality of care observed between 2001 and 2003 were as a result of GPs ‘gearing up’ for the introduction of the nGMS contract.

This is the first completed longitudinal study with data collected prospectively between the introduction of the NSF and nGMS contract, up until the present day. This study was powered by a large sample size, with all of the practices electing to take part in data collection throughout the study. The prospective design and data collection strengthens this study and we are confident the data represents accurately GP practices’ performance at each time point.

An important consideration when assessing the improvements in care is that, for many parameters, the recording was poor in the early years covered by this study. As there is likely to be preferential recording of patients with poor control, this may have resulted in artificially high values earlier in the study. As demonstrated in Table 1 there was a change in the characteristics of the population with diabetes over the course of the study, which included a small but significant increase in age and in time since diagnosis. It was not possible to link individual patients in the separate data collections, and as such it is impossible to quantify the flux of patients in the practices. The improvements observed reflect real improvements in patient care.

Conclusions
Our results confirm those of other studies that have demonstrated that the standard of diabetes care was already improving prior to the instigation of the nGMS contract (Seddon et al 2001; Campbell et al 2005; Doran et al 2006; Gray et al, 2006). The many changes to health care over the last 6 years have resulted in improved standards of care for people with diabetes in general practice. The NSF placed an emphasis on patient-centred care and developing a therapeutic relationship. The nGMS moved away from this towards a more biomedical model of care, focusing on easily quantifiable targets. Alongside this, the various guidelines that were introduced – including the NICE guidelines – equipped clinicians with evidence based methods for achieving the targets.

Acknowledgements
We are extremely grateful to the staff of the practices involved in this study for their longstanding support and cooperation. We would like to thank Chloe Todd for her help with data collection and training for the study, and Dr Ruth Milton.