Zack Cooper

Does Hospital Competition Save Lives? Evidence from the English NHS Patient Choice Reforms

Recent substantive reforms to the English National Health Service expanded patient choice and encouraged hospitals to compete within a market with fixed prices. This study investigates whether these reforms led to improvements in hospital quality. We use a difference-in-difference-style estimator to test whether hospital quality (measured using mortality from acute myocardial infarction) improved more quickly in more competitive markets after these reforms came into force in 2006. We find that after the reforms were implemented, mortality fell (i.e. quality improved) for patients living in more competitive markets. Our results suggest that hospital competition can lead to improvements in hospital quality.

In defence of our research on competition in England's National Health Service

2064 www.thelancet.com Vol 378 December 17/24/31, 2011 Submissions should be made via our electronic submission system at http://ees.elsevier.com/ thelancet/ strongest support for the causation hypothesis may be revealed”. Indeed, Bradford Hill also lavished praise on Snow, who examined the causes of cholera outbreaks in London in what is regarded as the fi rst use of diff erencein-diff erence regression. This is the same strategy we used to test the eff ect of competition. No study is perfect, which is why we have peer review and open science. However, the fact that three studies by separate research teams produced consistent results strongly fortifi es our collective fi ndings. More work surely needs to be done to understand the changes competition has brought about in England. However, the way forward should be to look objectively to see what is driving our fi ndings, rather than dismissing the results out of hand because they confl ict with prior beliefs.

Trends in Hospitalizations Among Medicare Survivors of Aortic Valve Replacement in the United States From 1999 to 2010

BACKGROUND Mortality rates after aortic valve replacement have declined, but little is known about the risk of hospitalization among survivors and how that has changed with time. METHODS Among Medicare patients who underwent aortic valve replacement from 1999 to 2010 and survived to 1 year, we assessed trends in 1-year hospitalization rates, mean cumulative length of stay (average number of hospitalization days per patient in the entire year), and adjusted annual Medicare payments per patient toward hospitalizations. We characterized hospitalizations by principal diagnosis and mean length of stay. RESULTS Among 1-year survivors of aortic valve replacement, 43% of patients were hospitalized within that year, of whom 44.5% were hospitalized within 30 days (19.2% for overall cohort). Hospitalization rates were higher for older (50.3% for >85 years), female (45.1%), and black (48.9%) patients. One-year hospitalization rate decreased from 44.2% (95% confidence interval, 43.5 to 44.8) in 1999 to 40.9% (95% confidence interval, 40.3 to 41.4) in 2010. Mean cumulative length of stay decreased from 4.8 days to 4.0 days (p < 0.05 for trend); annual Medicare payments per patient were unchanged (

Making competition work in the English NHS: the case for maintaining regulated prices

5,709 to

National Trends in Hospital Readmission Rates among Medicare Fee-for-Service Survivors of Mitral Valve Surgery, 1999–2010

5,737; p = 0.32 for trend). The three most common principal diagnoses in hospitalizations were heart failure (12.7%), arrhythmia (7.9%), and postoperative complications (4.4%). Mean length of stay declined from 6.0 days to 5.3 days (p < 0.05 for trend). CONCLUSIONS Among Medicare beneficiaries who survived 1 year after aortic valve replacement, 3 in 5 remained free of hospitalization; however, certain subgroups had higher rates of hospitalization. After the 30-day period, the hospitalization rate was similar to that of the general Medicare population. Hospitalization rates and cumulative days spent in hospital decreased with time.

Does hospital competition improve efficiency? An analysis of the recent market-based reforms to the English NHS

Over the last 20 years NHS policy-makers have increasingly relied on provider competition as a tool to drive improvements in hospital performance. This effort to promote competition began with the internal market in the 1990s, which separated the purchasers of care from the providers of care. It was followed in the mid-2000s by efforts to increase patient choice, expand the role of private sector providers, grant hospitals additional financial and managerial autonomy, and create a reimbursement system where money follows patients’ choices under fixed prices (Payment by Results). Now, the current coalition government is seeking to further expand the role of competition in the English NHS. The recent Health and Social Care Bill shifts purchasing power from primary care trusts (PCTs) to newly formed GP consortia, encourages private sector providers to play a more active role in health care provision and creates an economic regulator to manage competition in the NHS. Ultimately, the challenge that the current government is facing is deciding how to build on the market created by their predecessors, which researchers suggest is yielding positive results, within a much more cash constrained environment. –4 One key element of the current government’s reforms that has concerned us greatly is the proposal, outlined in both the Bill and the 2011–12 NHS Operating Framework, to shift from fixed prices to maximum prices, thereby allowing price competition. In response to growing concerns about their proposals, the government recently announced plans to amend the Bill to remove all reference to maximum prices. This was a sensible change in the legislation that the government should be given credit for making. We were concerned about introducing price competition in the NHS not because of an ideological opposition to the idea. Far from it – our concern is how best to incentivize providers to improve their quality and productivity. Indeed, we expect that in the years to come, it may be quite justified to introduce price competition in certain sectors of the health service. However, we are reticent about price competition in the Bill because the research evidence suggests that introducing price competition in environments where quality is difficult to measure and purchasers face significant pressure to constrain costs can harm clinical quality. The hospital competition literature has drawn a sharp distinction between markets where hospitals can compete on both price and quality, and markets where prices are fixed by a regulator and hospitals can only compete on quality. In markets which allow quality and price competition, theory predicts that hospital competition may either increase or decrease quality. Here, the response of providers depends very much on the preferences of patients, who often have hugely varying tastes and preferences about tradeoffs between costs and quality. In addition, in hospital markets, where quality is often vastly more difficult to measure than price, competition may harm quality as providers choose to differentiate themselves on the elements of care that purchasers of health care can easily observe (i.e. price) at the expense of those that they cannot (i.e. clinical quality). The theory that price competition can be harmful to clinical quality is supported by empirical research from the USA by Volpp and colleagues, who looked at the impact of price competition in New Jersey and found that it significantly increased mortality rates. Propper and colleagues looked at similar issues in the 1990s NHS internal market, which allowed price competition. Both UK studies found that during the internal market, greater competition was associated with lower clinical quality. In contrast, the theory about hospitals’ response to competition in markets with regulated prices is more straightforward. Theory predicts that faced with competition, as long as reimbursement rates are greater than hospitals’ marginal costs, hospitals will increase their quality in an effort to increase market share until their profits approach zero. Evidence from the USA broadly supports the idea that quality competition prompts hospitals to improve their performance. A range of studies, the most well know of which was by Kessler and McClellan, found that hospital competition prompts lower death rates and improvements in other aspects of clinical performance. Indeed, analyses of the impact of recent efforts to increase competition based on quality in the NHS have found similar results: hospitals exposed to greater potential competition as a result of the 2000s reforms was associated with improved quality and efficiency at a faster rate after the reforms took force in 2006. Likewise, Bloom and colleagues found that hospitals in England facing more potential competition was associated with better managed care and lower death rates. With this evidence in mind, while we support competition as a tool for improvement, we do not support reintroducing of price competition in to the NHS at this point. Given that the current market appears to be working positively, the challenge for the government is to build

Does competition improve public hospitals’ efficiency?: evidence from a quasi-experiment in the English National Health Service

Background Older patients who undergo mitral valve surgery (MVS) have high 1-year survival rates, but little is known about the experience of survivors. Our objective was to determine trends in 1-year hospital readmission rates and length of stay (LOS) in these individuals. Methods We included 100% of Medicare Fee-for-Service patients ≥65 years of age who underwent MVS between 1999–2010 and survived to 1 year (N = 146,877). We used proportional hazards regression to analyze the post-MVS 1-year readmission rate in each year, mean hospital LOS (after index admission), and readmission rates by subgroups (age, sex, race). Results The 1-year survival rate among patients undergoing MVS was 81.3%. Among survivors, 49.1% experienced a hospital readmission within 1 year. The post-MVS 1-year readmission rate declined from 1999–2010 (49.5% to 46.9%, P<0.01), and mean hospital LOS decreased from 6.2 to 5.3 (P<0.01). Readmission rates were highest in oldest patients, but declined in all age subgroups (65–74: 47.4% to 44.4%; 75–84: 51.4% to 49.2%, ≥85: 56.4% to 50.0%, all P<0.01). There were declines in women and men (women: 51.7% to 50.8%, P<0.01; men: 46.9% to 43.0%, P<0.01), and in whites and patients of other race, but not in blacks (whites: 49.0% to 46.2%, P<0.01; other: 55.0% to 48.9%, P<0.01; blacks: 58.1% to 59.0%, P = 0.18). Conclusions Among older adults surviving MVS to 1 year, slightly fewer than half experience a hospital readmission. There has been a modest decline in both the readmission rate and LOS over time, with worse outcomes in women and blacks.