Effect of Pay for Performance on Hypertension in the United Kingdom

Abstract

Commentary on Serumaga B, Ross-Degnan D, Avery AJ, et al. Effect of pay for performance on the management and outcomes of hypertension in the United Kingdom: interrupted time series study. BMJ. 2011;342:d108. Commentary on Serumaga B, Ross-Degnan D, Avery AJ, et al. Effect of pay for performance on the management and outcomes of hypertension in the United Kingdom: interrupted time series study. BMJ. 2011;342:d108. The Quality and Outcomes Framework (QOF) is a pay-for-performance (P4P) system in the United Kingdom intended to reward delivery of quality care in primary care practices and drive further improvements in care for patients.1The Health and Social Care Information CentreThe Quality and Outcomes Framework.http://www.ic.nhs.uk/statistics-and-data-collections/audits-and-performance/the-quality-and-outcomes-frameworkGoogle Scholar The QOF has a range of evidence-based national quality standards that cover 4 areas (clinical care, organizational factors, patient experience, and additional services) within chronic disease management. Each domain has measures of achievement (indicators) rewarded with points corresponding to the degree to which the standard is met. Currently, there is a maximum of 1,000 points, with the payout per point adjusted in terms of practice list size and relative prevalence of the covered chronic diseases (in 2009-2010, practices were paid an average of £126.77 [US $205.80] per point). These incentive payments represented 15% of practice income in 2009-2010. Overall achievement of quality standards has been high, with practices in England averaging >95% since the first year after inception.1The Health and Social Care Information CentreThe Quality and Outcomes Framework.http://www.ic.nhs.uk/statistics-and-data-collections/audits-and-performance/the-quality-and-outcomes-frameworkGoogle Scholar The clinical care domain comprises 70% of the QOF and is organized by disease category, which includes hypertension, coronary heart disease, heart failure, stroke, diabetes, and chronic kidney disease (CKD). One important feature is the establishment of disease registers, with payments to practices that are based on the number of patients on each register and subject to achievement thresholds that vary according to the indicator. The relationship of disease prevalence to payment is complex in that doubling disease prevalence on the register increases the relative payment by a factor of only √2. Points related to hypertension are listed in the Table 1; note that the blood pressure (BP) threshold is ≤150/90 mm Hg for most patients and lower in patients with diabetes (≤145/85 mm Hg) and CKD (≤140/85 mm Hg). Ultimately, the goal of the QOF and any other P4P system is to enhance the quality of care delivery; therefore, quality targets should be a proxy for improved long-term outcomes. Accordingly, in a study published in 2011 in the BMJ, Serumaga et al2Serumaga B. Ross-Degnan D. Avery A.J. et al.Effect of pay for performance on the management and outcomes of hypertension in the United Kingdom: interrupted time series study.BMJ. 2011; 342: d108Crossref PubMed Scopus (202) Google Scholar examined the impact of the UK QOF P4P system on both the treatment of and outcomes for hypertension in the United Kingdom.Table 1QOF Indicators, Points, and Payment Stages Relevant to HTNIndicatorPointsPayment Stages (%)BP 1: The practice can produce a register of patients with established HTN6—BP 4: The percentage of patients with HTN for whom there is a record of the BP in the previous 9 mo1840-90BP 5: The percentage of patients with HTN for whom the last BP was ≤150/90 mm Hg, for measurements in the previous 9 mo5740-70CHD 1: The practice can produce a register of patients with CHD4—CHD 5: The percentage of patients with CHD whose notes have a record of BP in the previous 15 mo740-90CHD 6: The percentage of patients with CHD for whom the last BP was ≤150/90 mm Hg, for measurements in the previous 15 mo1740-70HF 1: The practice can produce a register of patients with heart failure4—Stroke 1: The practice can produce a register of patients with stroke or TIA2—Stroke 5: The percentage of patients with TIA or stroke who have a record of BP in the notes in the preceding 15 mo240-90Stroke 6: The percentage of patients with a history of TIA or stroke for whom the last BP reading was ≤150/90 mm Hg, for measurements in the previous 15 mo540-70CKD 1: The practice can produce a register of patients aged ≥18 y with CKDaCKD stages 3-5 according to the US National Kidney Foundation's KDOQI (Kidney Disease Outcomes Quality Initiative) definitions.6—CKD 2: The percentage of patients on the CKD register whose notes have a record of BP in the previous 15 mo640-90CKD 3: The percentage of patients on the CKD register for whom the last BP was ≤140/85 mm Hg, for measurements in the previous 15 mo1140-70DM 11: The percentage of patients with DM who have a record of the BP in the previous 15 mo340-90DM 12: The percentage of patients with DM for whom the last BP was ≤145/85 mm Hg1840-60Records 11: The BP of patients aged ≥45 y is recorded in the preceding 5 y for ≥65% of patients10—Records 17: The BP of patients aged ≥45 y is recorded in the preceding 5 y for ≥80% of patients5—Note: The QOF measures achievement against a range of evidence-based indicators, with points and payments awarded according to the level of achievement.Abbreviations: BP, blood pressure; CHD, coronary heart disease; CKD, chronic kidney disease; DM, diabetes mellitus; HTN, hypertension; QOF, Quality and Outcomes Framework; TIA, transient ischemic attack.a CKD stages 3-5 according to the US National Kidney Foundation's KDOQI (Kidney Disease Outcomes Quality Initiative) definitions. Open table in a new tab Note: The QOF measures achievement against a range of evidence-based indicators, with points and payments awarded according to the level of achievement. Abbreviations: BP, blood pressure; CHD, coronary heart disease; CKD, chronic kidney disease; DM, diabetes mellitus; HTN, hypertension; QOF, Quality and Outcomes Framework; TIA, transient ischemic attack. Serumaga et al2Serumaga B. Ross-Degnan D. Avery A.J. et al.Effect of pay for performance on the management and outcomes of hypertension in the United Kingdom: interrupted time series study.BMJ. 2011; 342: d108Crossref PubMed Scopus (202) Google Scholar examined data from 470,725 primary care records, obtained from the UK THIN (The Health Improvement Network) database, to evaluate the impact of the QOF P4P system on the care and outcomes of people given a diagnosis of hypertension from January 2000 to August 2007. Modifications in management (frequency of therapy initiation and monitoring and volume of prescribing), intermediate outcomes (centiles of BP and extent of hypertension control), and clinical outcomes (all-cause mortality, myocardial infarction, heart failure, stroke, and kidney failure) were measured in the 3 years before and up to 4 years after enactment of P4P. Analytic methods included interrupted time series analysis controlling for baseline level and trend and segmented survival analysis. Subgroups of newly treated (treatment started 6-12 months pre-QOF; n = 103,009) and treatment-experienced patients (treatment started ≥3 years pre-QOF; n = 104,754) were followed up for up to 45 and 80 months, including 40 months after initiation of the QOF for both subgroups, respectively. Metrics included the number of patients with BP <150/90 mm Hg as a fraction of all those with BP readings every quarter, proportion of patients with a BP measure each month, and number of antihypertensive drugs prescribed per patient each month. Overall, there was no apparent affect of the P4P program. There were no significant changes in systolic and diastolic BPs or rate of controlled BP (∼70%) during the study. BP monitoring increased slightly from 45.6% at the start of observation by 0.15% per month, but with no change after implementation of P4P. Treatment intensified in the 4-year baseline period, with 0.22% (P < 0.001) fewer people receiving either no drugs or only one drug and a simultaneous increase in those receiving either 2 (0.02%; P < 0.001) or ≥3 drugs (0.19%; P < 0.001). No change in treatment intensity was seen after P4P implementation. Cumulative incidence rates of any adverse clinical outcome (myocardial infarction, stroke, or heart or kidney failure) or all-cause mortality for both patient subgroups (newly treated and treatment experienced) increased linearly both before and after P4P with no change in level or trend temporally related to P4P. The present study has several limitations that may be specific to this data source. First, there are errors in the choice of diagnostic (Read) codes used to define hypertension in this study. The THIN database, from which patient data were obtained, draws its codes from the Read, version 2, 5-Byte hierarchy3de Lusignan S. Codes, classifications, terminologies and nomenclatures: definition, development and application in practice.Inform Prim Care. 2005; 13: 65-70PubMed Google Scholar rather than the Clinical Terms, version 3, as stated in the report. The code lists used are limited and do not accurately match the code list used for QOF (all available online1The Health and Social Care Information CentreThe Quality and Outcomes Framework.http://www.ic.nhs.uk/statistics-and-data-collections/audits-and-performance/the-quality-and-outcomes-frameworkGoogle Scholar), thus confounding measures of numbers of patients on disease registers. For example, codes for hypertension used in this study included G20 (essential hypertension), but not G2 (hypertensive disease). In a convenience sample of 2008 data available from the QICKD (Quality Improvement in Chronic Kidney Disease) trial, 16% (17,392 of 108,698) of people have a G2 instead of a G20 code.4de Lusignan S. Gallagher H. Chan T. et al.The QICKD Study protocol: a cluster randomised trial to compare quality improvement interventions to lower systolic BP in chronic kidney disease (CKD) in primary care.Implement Sci. 2009; 4: 39Crossref PubMed Scopus (39) Google Scholar There are similar inconsistencies in codes included for stroke and heart failure. In sum, the choice of codes means that the population studied is not identical to the P4P population; therefore, caution is recommended in interpretation of results. Second, although the standard of care for hypertension in UK patients was improving before implementation of P4P, the trends in BP control found by Serumaga et al2Serumaga B. Ross-Degnan D. Avery A.J. et al.Effect of pay for performance on the management and outcomes of hypertension in the United Kingdom: interrupted time series study.BMJ. 2011; 342: d108Crossref PubMed Scopus (202) Google Scholar are at variance with other data. The Health Survey for England (conducted each year to measure health and health-related habits in children and adults residing in private households) has shown continued improvement in the management of hypertension since 1994.5The Health and Social Care Information CentreHealth Survey for England—2009: Trend tables.http://www.ic.nhs.uk/statistics-and-data-collections/health-and-lifestyles-related-surveys/health-survey-for-england/health-survey-for-england--2009-trend-tablesGoogle Scholar Similarly, evaluation of improvements in hypertension management from 2003 to 2006, based on random samples representative of UK noninstitutionalized adults (n = 8,834 in 2003 and n = 7,478 in 2006) showed meaningful and significant increases in rates of awareness, therapy, and control, especially in women.6Falaschetti E. Chaudhury M. Mindell J. Poulter N. Continued improvement in hypertension management in England: results from the Health Survey for England 2006.Hypertension. 2009; 53: 480-486Crossref PubMed Scopus (187) Google Scholar Analysis of data from 236,467 patients in another large UK primary care database (the Doctors Independent Network [DIN-LINK] GP database) in 2000-2005 also found improvements in hypertension control. Recorded true systolic BP >150 mm Hg decreased from 36% in 2000-2001 to 23% in 2004-2005 and treatment intensity increased.7Carey I.M. Nightingale C.M. DeWilde S. Harris T. Whincup P.H. Cook D.G. Blood pressure recording bias during a period when the Quality and Outcomes Framework was introduced.J Hum Hypertens. 2009; 23: 764-770Crossref PubMed Scopus (23) Google Scholar The investigators of both studies concluded that the pattern of decreasing BP levels in UK primary care patients has persisted beyond the initiation of P4P. Finally, in direct contrast to Serumaga et al,2Serumaga B. Ross-Degnan D. Avery A.J. et al.Effect of pay for performance on the management and outcomes of hypertension in the United Kingdom: interrupted time series study.BMJ. 2011; 342: d108Crossref PubMed Scopus (202) Google Scholar analysis of QOF data from April 2004 to March 2007 for all UK practices found substantial improvements in BP monitoring and control after the introduction of P4P.8Ashworth M. Medina J. Morgan M. Effect of social deprivation on blood pressure monitoring and control in England: a survey of data from the Quality and Outcomes Framework.BMJ. 2008; 337: a2030Crossref PubMed Scopus (88) Google Scholar This analysis addressed BP indicators in patients on the hypertension, coronary heart disease, stroke, diabetes, and CKD registers and was aimed at determining the effect of social deprivation. Although target achievement at the start of the study was less in the most versus least deprived communities, better BP target attainment in the 5 disease areas was associated with almost total abrogation of the achievement gap separating these 2 community types. This effect was attributed to P4P. A cohort study examining data from the General Practice Research Database found that the percentage of patients with diabetes who reached the BP target exceeded predicted levels by 3.1% (95% confidence interval, 1.1%-5.1%; P < 0.001) after implementation of the QOF compared with that predicted by the preintervention trend.9Millett C. Bottle A. Ng A. et al.Pay for performance and the quality of diabetes management in individuals with and without co-morbid medical conditions.J R Soc Med. 2009; 102: 369-377Crossref PubMed Scopus (42) Google Scholar Campbell et al10Campbell S. Reeves D. Kontopantelis E. Middleton E. Sibbald B. Roland M. Quality of primary care in England with the introduction of pay for performance.N Engl J Med. 2007; 357: 181-190Crossref PubMed Scopus (318) Google Scholar reported an increase in rate of improvement in the quality of care for patients with diabetes in the United Kingdom after P4P, but not for coronary heart disease. In a 10-year longitudinal study of patients with coronary heart disease spanning implementation of P4P (from 1998-2007), the percentage of patients who had a BP measurement recorded increased from 33.2% to 93.9%.11Murray J. Saxena S. Millet C. Curcin V. de Lusignan S. Majeed A. Reductions in risk factors for secondary prevention of coronary heart disease by ethnic group in South-West London: 10-year longitudinal study (1998-2007).Fam Pract. 2010; 27: 430-438Crossref PubMed Scopus (17) Google Scholar Mean BP decreased from 140/80 to 133/74 mm Hg (P < 0.001). Finally, a time series analysis of 3.4 million patients from 498 practices in England during 2001-2006 showed increased recorded prevalences of hypertension, coronary heart disease, stroke, diabetes, and CKD during the 5-year study period.12Hippisley-Cox J. Vinogradova Y. Coupland C. Time series analysis for selected clinical indicators from the Quality and Outcomes Framework 2001-2006 Final report for the information centre, version 1.1, R22 HSCIC. 2007.http://www.qresearch.org/Public_Documents/Time%20Series%20Analysis%20for%20selected%20clinical.pdfGoogle Scholar The relative increase in percentage of patients with controlled BP levels was at least 50% in all 5 disease areas. P4P represents one of the main tools for driving quality improvement programs in health care, but a major challenge has been the development and implementation of measures of performance. Hypertension is a key risk factor for cardiovascular disease and predicts the development and progression of CKD. For the United Kingdom, it has been calculated that if systolic BP of all hypertensive adults could be controlled to 140 mm Hg, there would be decreases of 28%-44% for stroke and 20%-35% for coronary heart disease, depending on age.13He F.J. MacGregor G.A. Cost of poor blood pressure control in the UK: 62,000 unnecessary deaths per year.J Hum Hypertens. 2003; 17: 455-457Crossref PubMed Scopus (85) Google Scholar These important outcomes, although common in the population as a whole, are not common enough at the level of individual providers, are too multifactorial, and often manifest too far in the future to provide immediate gratification to motivate quality improvement good behavior. Accordingly, intermediate metrics, such as recording and control of BP, that can be measured easily, are influenced relatively easily, and are part of the causal pathway are used. However, consistent evidence for a benefit from implementation of P4P in hypertension in the United Kingdom is clearly lacking. Aside from coding issues, it is possible that the method used by Serumaga et al2Serumaga B. Ross-Degnan D. Avery A.J. et al.Effect of pay for performance on the management and outcomes of hypertension in the United Kingdom: interrupted time series study.BMJ. 2011; 342: d108Crossref PubMed Scopus (202) Google Scholar in assessing the effectiveness of QOF was flawed and incremental changes toward achievement of a target are not of equal value. Stated differently, the closer one gets to a target, the greater the treatment challenge to achieve that target, and without P4P, the trend of continued improvement may have leveled off. Nevertheless, Serumaga et al2Serumaga B. Ross-Degnan D. Avery A.J. et al.Effect of pay for performance on the management and outcomes of hypertension in the United Kingdom: interrupted time series study.BMJ. 2011; 342: d108Crossref PubMed Scopus (202) Google Scholar suggest that the bar for P4P targets for hypertension alone may be set too low, possibly explaining why the more demanding target in diabetes appears to have been consistently influenced positively by P4P. Aligning payment to performance involves setting a “goal gradient” between baseline performance and goal performance. Too steep a gradient discourages effort because the goal is seen as unattainable; too shallow a gradient, although encouraging greater effort to achieve an attainable goal, then promotes performance decay when the goal is achieved.14Campbell S.M. Reeves D. Kontopantelis E. Sibbald B. Roland M. Effects of pay for performance on the quality of primary care in England.N Engl J Med. 2009; 361: 368-378Crossref PubMed Scopus (501) Google Scholar Finally and critically, to achieve sustained improvement in population BP control, future developments in P4P need to de-emphasize process indicators, which reward practices for carrying out tasks such as BP checks in favor of a focus on outcome measures, such as optimal control of BP. Drs Stevens and de Lusignan are nephrologist and general practitioner advisors for the CKD QOF indicators, respectively, and Dr Stevens is a member of the UK Department of Health's Renal Advisory Group. Financial Disclosure: Dr de Lusignan has received lecture fees from Pfizer. The remaining authors declare that they have no relevant financial interests.