Do variations in practice core income buy improved performance in Bristol 2011-12?


I have already shown that there was a twofold variation in core practice income per patient per year in Bristol in 2011-12. I have also been unable to show that this variation is related to obvious and available measures of patient need. In fact there are some worrying negative correlations when need and income are examined. In this report I will look at common available measures of GP practice performance and see if I can detect any evidence that increased resourcing is cost effective.


Immunisations

All 4 of these immunisation targets show no positive association with practice core income. There is a ceiling effect here, with a large number of practices achieving high rates of immunisation. Note a tight group at the £60 mark, the GMS practices receive a payment for this target. The outliers do not seem to be income related.

 
Figure 1, Pediacel 5 in one vaccine(DTPPH) rates, children aged 2 years, R= -0.10

 

Figure 2 : meningitis vaccine aged 2, R=-0.14



Figure 3: MMR at 2 years, R=-0.06

Figure 4 : 5 year booster rate , R= -0.13

Cervical Smears
No evidence that higher practice income improves cervical screening. The 2 lowest screening rates are from an east Bristol practice(with a large Asian population) and Student Health. Again a very tight group at £60, largely GMS practices who obtain a target payment for cervical cytology.

Figure 5: cervical smear uptake, R=-0-015

Patient experience
There are various measures of patient experience available from NHS patient surveys. The data shown here refer to overall satisfaction with the practice, good or better. Despite quite a narrow outcome range, there is a small (non-significant) positive association between income and the percentage of patients reporting a good experience. Perhaps this is because some higher earning practices can offer better telephone response and earlier and more flexible appointments?

Figure 6, Patient experience good or better, R=0.077

Hypertensive control, BP< 150/90
At last some evidence that better paid practices can deliver better outcome? Not a significant correlation, this is a measure incentivised via QOF, and we know from previous results that higher earners generally do well with QOF.

Figure 7 : Hypertensive control, R= 0.16

Standardised Mortality Ratio(SMR)
SMR would ideally be a performance measure, with good practices prolonging life expectation and thus having low SMRS. There are some very large big differences in SMR in Bristol, from 69 in Westbury-on- Trym to 140 in Hartcliffe, reflecting wide social differences. Most observers would tend to prefer resources to be targeted at areas with high SMRs. That does not seem to be happening. As with a lot of the data presented, there is disappointingly little association.

Figure 8: SMR ,  R= -0.14

Similar data for SMR aged <65, R=-0.12, not shown.

Interpretation
There is not much evidence to show that variation in practice core income is reflected in performance using these fairly basic measures. We know that QOF can incentivise GPs to deliver measurable high performance in those QOF areas : QOF accounts for 14% of GP income in Bristol. A large proportion of GP core income varies from £60-120 per patient with little measurable difference in service performance. Some measures have small but worryingly negative associations.

click here for PDF