This site is intended for health professionals only

GP practices urged to take ‘supermarket-style’ approach to boost clinical outcomes

GP practices need to concentrate on the ‘bigger picture’ and improve the experience of patients in their surgeries as this can affect their outcomes, claims a leading thinktank.

The King’s Fund paper said patients at practices with better patient satisfaction scores – particularly for access - were more likely to attend their appointments, take their medicines and have better outcomes.

Their analysis – published this week – compared the performance of over 8,000 GP practices and they showed a correlation between practices scoring highly on delivering a good service, especially patient access, and their performance on measures of clinical quality.

The researchers also found practices which performed poorly on both patient experience and clinical outcomes are mainly based in London, were in areas of higher deprivation and also tended to have fewer GPs.

They said practices should learn lessons from supermarkets and department stores in treating patients as ‘customers’.

Researchers analysed QOF scores from 2010/11 for nine conditions, including diabetes, hypertension and stroke, and compared them with GP patient survey scores for measures such as access, cleanliness, information and involvement.

They found a significant statistical correlation with some of the patient experience scores and QOF achievement, but access scores were strongly and consistently correlated with all aspects of patient experience (with p values of less than or equal to zero).

In terms of overall satisfaction, QOF process measures for hypertension, CHD, COPD, depression and stroke had p values of zero. Process measures for asthma, dementia, smoking and heart failure were not signficantly correlated with overall patient satisfaction. QOF outcomes measures for hypertension, CHD, diabetes and stroke all had p values of less than 0.01.

The authors concluded, although they could not prove a ‘causal link’, patient experience was important for the outcomes of patients.

The report said: ‘Our analysis suggests that a patient’s ease of access to their practice and preferred GP could affect their quality of care and outcomes – for example, through its impact on attendance rates, continuity of care, communication and engagement with clinical staff, compliance and adherence with treatment, and out-of-hours access.

They added: ‘We urge staff in general practice to consider how they perform in terms of the patient experience alongside how they perform on the quality of clinical care, as it is this bigger picture that offers the greatest potential for driving further improvements in the overall quality of care.’

Study leader Dr Veena Raleigh, a senior fellow at the King’s Fund, said GP practices should compare their services with the customer service at supermarkets and department stores.

She said: ‘I think the analogy with the commercial sector is Sainsbury and John Lewis like to have happy customers.’

But Dr Osman Bhatti, a GP in Tower Hamlets, east London, and a member of Newham LMC said the study was undermined by the ‘skewed ‘ results of the patient survey.

He said: ‘Satisfaction in more deprived areas is often tended to be skewed down as most of those that reply are the ones who are going to comment negatively and often those who have no issue and feel they are well looked after, often don’t respond to such surveys.’

Dr Andrew Mimnagh, a GP in Liverpool and chair of Sefton LMC said: ‘The problem with patient outcome surveys is that a good bedside manner can hide a multitude of sins.’

Health minister, Earl Howe said the changes they were recommending for the GP contract would improve the problems identified in the report.

He said: ‘We want to tackle variation in quality of care and services by moving to a fairer system for rewarding GPs.

‘Payments should better reflect the number of patients on GPs lists and be appropriately weighted for factors such as age and deprivation.’

QOF vs. Patient satisfaction
*** p<=0.000
** p<=0.01
NS = non-significant correlation