Thesis Thursday: Sarah Zheng

On the third Thursday of every month, we speak to a recent graduate about their thesis and their studies. This month’s guest is Dr Sarah Zheng who has a PhD from Boston University. If you would like to suggest a candidate for an upcoming Thesis Thursday, get in touch.

Title
Design for performance: studies on cost and quality in U.S. health care
Supervisors
Z. Justin Ren, Kimberley H. Geissler, Janelle Heineke, Anita Tucker
Repository link
https://open.bu.edu/handle/2144/23312

In the context of your PhD research, what does ‘design for performance’ mean?

“Design for performance” is a further step in managing healthcare from “pay for performance”, on which there has been decades of attention paid among practitioners and academics. Despite the long effort on “pay for performance”, the core challenge remains how to properly incentivize patients, clinicians and staff to align their behaviors with optimal, safe and cost-effective, patient-centric care. This dissertation suggests an important set of issues to consider around “design for performance” at the system and process levels.

At the system level, under what conditions does cost-sharing lead to lower total costs without reducing quality of care? Previous literature has studied contract theory and mechanism design in varied industry settings (Guajardo et al. 2012), yet very few are studied in the healthcare domain where insurance plans are offered to patients under different contract arrangements. It remains unclear whether certain contract design at such settings may lead to desired outcomes (e.g., low healthcare spending). At the process level, under what conditions and to what extent does excellent internal supply operations result in superior hospital performance? Industrial studies suggest that reliable, efficient internal supply chains that are integrated with production yield better financial and quality performance for manufacturing companies (Droge et al. 2004, Flynn et al. 2010). However, there is scant quantitative research on the impact of support departments in hospitals (Tucker et al. 2008, Fredendall et al. 2009). Studies are needed to understand the extent to which support departments impact patient care outcomes, such as adverse events.

How was quality captured in the data that you used in your analyses?

In Chapter 3 of my thesis, I studied the impact of internal service quality on one particular quality performance metric: adverse events. Specifically, it is a rate variable that is calculated by the sum of adverse events (i.e., patient falls with injury and pressure ulcers) on the unit that month divided by the number of patient days on the unit that month, which is then multiplied by 1,000. The hospital collects these data monthly. The adverse event data come from both patient record reviews and incident reports in the hospital’s safety reporting system, as is typical of this type of data (Lake and Cheung 2006). The error event data are audited internally as well as reported to CMS (Zheng et al. 2017).

This is a unique opportunity to study quality as most healthcare operations research has relied on publicly available, hospital-level quality data, such as patient mortality (e.g., Senot et al. 2015, KC and Terwiesch 2011)—which is a blunt measure of quality—or process of care measures (e.g., Boyer et al. 2012, Gardner et al. 2015, Senot et al. 2015), which have been criticized in the healthcare literature for their weak connection to clinical outcomes (Patterson et al. 2010).

You complemented your quantitative analysis with some qualitative interviews – was this a valuable exercise?

Yes, definitely. To understand further the role patients (and the patient-physician dyad) play in deciding the usage of imaging studies, I conducted in-depth conversations with both physicians and patients. Specifically, I interviewed three physicians (i.e., hospitalist, primary care provider) and one imaging technician with the average conversation time of 70 mins. I also interviewed six patients with the average conversation time of 20 mins.

I found that patients did play a role in deciding the usage of imaging studies in the way that high-deductible health plan (HDHP) patients are less likely to demand imaging studies than non-HDHP patients. However, as patients cannot distinguish low-value care from high-value care, HDHP patients avoid patient care in general. This is consistent with previous literature on patient cost-sharing and HDHPs where patients indiscriminately reduce medical care (Hibbard et al. 2008, Lohr et al. 1986). It further suggests that HDHP may be a blunt instrument, reducing all diagnostic imaging, rather than helping physicians and patients choose high-value imaging.

Did any of your findings about high-deductible health plans stand out as different from previous studies?

I wouldn’t say different but more like complementary. Previous studies found HDHPs have different impacts depending on the site and type of care (Haviland et al. 2015, Wharam et al. 2013, Bundorf 2012, Nair et al. 2009, Waters et al. 2011, Hibbard et al. 2008, Busch et al. 2006, Rowe et al. 2008, Parente et al. 2004). By explicitly testing associations between HDHP enrollment and diagnostic imaging, we provide a more complete picture for policymakers in making guidelines related to HDHP plans. Our results suggest that increases in HDHP enrollment may contribute to a slow in the growth of diagnostic imaging utilization. However, increased cost-sharing may not allow patients to differentiate between high-value and low-value utilization, and better patient awareness and education should be a crucial part of any reductions in diagnostic imaging utilization (Zheng et al. 2016).

‘Internal service quality’ is a term that doesn’t often appear in health economics journals – should researchers be dedicating more attention to this?

Yes. In our study we find improved internal service quality to be a particularly novel driver of reduced adverse events because it is not obvious a priori that support departments—most of which are not clinical in nature—could have a significant impact on clinical outcomes. In particular, we find that improving the overall average internal service quality received by a nursing unit by 0.1 on a 5-point scale is associated with a 38% reduction in adverse events per nursing unit, which has roughly the same benefit for reducing adverse events as increasing staffing on that unit by nearly one full-time equivalent nurse. In the hospital that we study, the average salary of a support service technician is lower than the average salary of a nurse. Thus, hospitals might be able to improve quality of care at a lower cost by increasing support staff to relieve the burden on nurses (Zheng et al. 2017). More studies are needed in this area to explore further internal service quality as a viable and cost-effective means to improve clinical performance.

Thesis Thursday: Francesco Longo

On the third Thursday of every month, we speak to a recent graduate about their thesis and their studies. This month’s guest is Dr Francesco Longo who has a PhD from the University of York. If you would like to suggest a candidate for an upcoming Thesis Thursday, get in touch.

Title
Essays on hospital performance in England
Supervisor
Luigi Siciliani
Repository link
http://etheses.whiterose.ac.uk/18975/

What do you mean by ‘hospital performance’, and how is it measured?

The concept of performance in the healthcare sector covers a number of dimensions including responsiveness, affordability, accessibility, quality, and efficiency. A PhD does not normally provide enough time to investigate all these aspects and, hence, my thesis mostly focuses on quality and efficiency in the hospital sector. The concept of quality or efficiency of a hospital is also surprisingly broad and, as a consequence, perfect quality and efficiency measures do not exist. For example, mortality and readmissions are good clinical quality measures but the majority of hospital patients do not die and are not readmitted. How well does the hospital treat these patients? Similarly for efficiency: knowing that a hospital is more efficient because it now has lower costs is essential, but how is that hospital actually reducing costs? My thesis tries to answer also these questions by analysing various quality and efficiency indicators. For example, Chapter 3 uses quality measures such as overall and condition-specific mortality, overall readmissions, and patient-reported outcomes for hip replacement. It also uses efficiency indicators such as bed occupancy, cancelled elective operations, and cost indexes. Chapter 4 analyses additional efficiency indicators, such as admissions per bed, the proportion of day cases, and proportion of untouched meals.

You dedicated a lot of effort to comparing specialist and general hospitals. Why is this important?

The first part of my thesis focuses on specialisation, i.e. an organisational form which is supposed to generate greater efficiency, quality, and responsiveness but not necessarily lower costs. Some evidence from the US suggests that orthopaedic and surgical hospitals had 20 percent higher inpatient costs because of, for example, higher staffing levels and better quality of care. In the English NHS, specialist hospitals play an important role because they deliver high proportions of specialised services, commonly low-volume but high-cost treatments for patients with complex and rare conditions. Specialist hospitals, therefore, allow the achievement of a critical mass of clinical expertise to ensure patients receive specialised treatments that produce better health outcomes. More precisely, my thesis focuses on specialist orthopaedic hospitals which, for instance, provide 90% of bone and soft tissue sarcomas surgeries, and 50% of scoliosis treatments. It is therefore important to investigate the financial viability of specialist orthopaedic hospitals relative to general hospitals that undertake similar activities, under the current payment system. The thesis implements weighted least square regressions to compare profit margins between specialist and general hospitals. Specialist orthopaedic hospitals are found to have lower profit margins, which are explained by patient characteristics such as age and severity. This means that, under the current payment system, providers that generally attract more complex patients such as specialist orthopaedic hospitals may be financially disadvantaged.

In what way is your analysis of competition in the NHS distinct from that of previous studies?

The second part of my thesis investigates the effect of competition on quality and efficiency under two different perspectives. First, it explores whether under competitive pressures neighbouring hospitals strategically interact in quality and efficiency, i.e. whether a hospital’s quality and efficiency respond to neighbouring hospitals’ quality and efficiency. Previous studies on English hospitals analyse strategic interactions only in quality and they employ cross-sectional spatial econometric models. Instead, my thesis uses panel spatial econometric models and a cross-sectional IV model in order to make causal statements about the existence of strategic interactions among rival hospitals. Second, the thesis examines the direct effect of hospital competition on efficiency. The previous empirical literature has studied this topic by focusing on two measures of efficiency such as unit costs and length of stay measured at the aggregate level or for a specific procedure (hip and knee replacement). My thesis provides a richer analysis by examining a wider range of efficiency dimensions. It combines a difference-in-difference strategy, commonly used in the literature, with Seemingly Unrelated Regression models to estimate the effect of competition on efficiency and enhance the precision of the estimates. Moreover, the thesis tests whether the effect of competition varies for more or less efficient hospitals using an unconditional quantile regression approach.

Where should researchers turn next to help policymakers understand hospital performance?

Hospitals are complex organisations and the idea of performance within this context is multifaceted. Even when we focus on a single performance dimension such as quality or efficiency, it is difficult to identify a measure that could work as a comprehensive proxy. It is therefore important to decompose as much as possible the analysis by exploring indicators capturing complementary aspects of the performance dimension of interest. This practice is likely to generate findings that are readily interpretable by policymakers. For instance, some results from my thesis suggest that hospital competition improves efficiency by reducing admissions per bed. Such an effect is driven by a reduction in the number of beds rather than an increase in the number of admissions. In addition, competition improves efficiency by pushing hospitals to increase the proportion of day cases. These findings may help to explain why other studies in the literature find that competition decreases length of stay: hospitals may replace elective patients, who occupy hospital beds for one or more nights, with day case patients, who are instead likely to be discharged the same day of admission.

Sam Watson’s journal round-up for 21st August 2017

Every Monday our authors provide a round-up of some of the most recently published peer reviewed articles from the field. We don’t cover everything, or even what’s most important – just a few papers that have interested the author. Visit our Resources page for links to more journals or follow the HealthEconBot. If you’d like to write one of our weekly journal round-ups, get in touch.

Multidimensional performance assessment of public sector organisations using dominance criteria. Health Economics [RePEcPublished 18th August 2017

The empirical assessment of the performance or quality of public organisations such as health care providers is an interesting and oft-tackled problem. Despite the development of sophisticated methods in a large and growing literature, public bodies continue to use demonstrably inaccurate or misleading statistics such as the standardised mortality ratio (SMR). Apart from the issue that these statistics may not be very well correlated with underlying quality, organisations may improve on a given measure by sacrificing their performance on another outcome valued by different stakeholders. One example from a few years ago showed how hospital rankings based upon SMRs shifted significantly if one took into account readmission rates and their correlation with SMRs. This paper advances this thinking a step further by considering multiple outcomes potentially valued by stakeholders and using dominance criteria to compare hospitals. A hospital dominates another if it performs at least as well or better across all outcomes. Importantly, correlation between these measures is captured in a multilevel model. I am an advocate of this type of approach, that is, the use of multilevel models to combine information across multiple ‘dimensions’ of quality. Indeed, my only real criticism would be that it doesn’t go far enough! The multivariate normal model used in the paper assumes a linear relationship between outcomes in their conditional distributions. Similarly, an instrumental variable model is also used (using the now routine distance-to-health-facility instrumental variable) that also assumes a linear relationship between outcomes and ‘unobserved heterogeneity’. The complex behaviour of health care providers may well suggest these assumptions do not hold – for example, failing institutions may well show poor performance across the board, while other facilities are able to trade-off outcomes with one another. This would suggest a non-linear relationship. I’m also finding it hard to get my head around the IV model: in particular what the covariance matrix for the whole model is and if correlations are permitted in these models at multiple levels as well. Nevertheless, it’s an interesting take on the performance question, but my faith that decent methods like this will be used in practice continues to wane as organisations such as Dr Foster still dominate quality monitoring.

A simultaneous equation approach to estimating HIV prevalence with nonignorable missing responses. Journal of the American Statistical Association [RePEcPublished August 2017

Non-response is a problem encountered more often than not in survey based data collection. For many public health applications though, surveys are the primary way of determining the prevalence and distribution of disease, knowledge of which is required for effective public health policy. Methods such as multiple imputation can be used in the face of missing data, but this requires an assumption that the data are missing at random. For disease surveys this is unlikely to be true. For example, the stigma around HIV may make many people choose not to respond to an HIV survey, thus leading to a situation where data are missing not at random. This paper tackles the question of estimating HIV prevalence in the face of informative non-response. Most economists are familiar with the Heckman selection model, which is a way of correcting for sample selection bias. The Heckman model is typically estimated or viewed as a control function approach in which the residuals from a selection model are used in a model for the outcome of interest to control for unobserved heterogeneity. An alternative way of representing this model is as copula between a survey response variable and the response variable itself. This representation is more flexible and permits a variety of models for both selection and outcomes. This paper includes spatial effects (given the nature of disease transmission) not only in the selection and outcomes models, but also in the model for the mixing parameter between the two marginal distributions, which allows the degree of informative non-response to differ by location and be correlated over space. The instrumental variable used is the identity of the interviewer since different interviewers are expected to be more or less successful at collecting data independent of the status of the individual being interviewed.

Clustered multistate models with observation level random effects, mover–stayer effects and dynamic covariates: modelling transition intensities and sojourn times in a study of psoriatic arthritis. Journal of the Royal Statistical Society: Series C [ArXiv] Published 25th July 2017

Modelling the progression of disease accurately is important for economic evaluation. A delicate balance between bias and variance should be sought: a model too simple will be wrong for most people, a model too complex will be too uncertain. A huge range of models therefore exists from ‘simple’ decision trees to ‘complex’ patient-level simulations. A popular choice are multistate models, such as Markov models, which provide a convenient framework for examining the evolution of stochastic processes and systems. A common feature of such models is the Markov property, which is that the probability of moving to a given state is independent of what has happened previously. This can be relaxed by adding covariates to model transition properties that capture event history or other salient features. This paper provides a neat example of extending this approach further in the case of arthritis. The development of arthritic damage in a hand joint can be described by a multistate model, but there are obviously multiple joints in one hand. What is more, the outcomes in any one joint are not likely to be independent of one another. This paper describes a multilevel model of transition probabilities for multiple correlated processes along with other extensions like dynamic covariates and different mover-stayer probabilities.

Credits