Discovery Health reviewed 8 years of data from private hospital “patient satisfaction surveys”. They say patients with low scores are more likely to have an unplanned hospital readmission in the 30 days after hospital discharge and that their study identifies responsible gaps in hospital care.
Are they right?
I want to discuss these conclusions and to draw attention to four aspects of this study:
- exactly what Discovery measured and how
- factors affecting patient experience and hospital readmission
- low survey response rates, and
- the improvement in scores seen over recent years.
Image: Google Gemini
1. The terms “patient satisfaction” and “patient experience” are often used interchangeably, but they are different concepts.
Although unacknowledged in the article, Discovery uses the HCAHPS (Hospital Consumer Assessment of Healthcare Providers and Systems) survey from the United States Agency for Healthcare Research and Quality (AHRQ). I know because in my past role as a consultant to Discovery, I introduced and helped implement it. HCAHPS measures patient experience, not patient satisfaction, during a hospital admission. The distinction is important.
Patient satisfaction is a subjective assessment reflecting a patient’s personal evaluation, feelings, and expectations about healthcare services received. It is influenced by previous experiences, personal and culturally based health beliefs, outcomes of care, and individual expectations, which can vary widely. Satisfaction levels can differ significantly among patients receiving the same level of care, as personal feelings and expectations are subjective and influenced by factors outside the direct control of healthcare professionals.
Patient experience, in contrast, encompasses all interactions and aspects of care that a patient encounters within the healthcare system. Assessment of patient experience focuses on specific, concrete, and more objective measures of what actually happens during care, such as communication with clinicians, responsiveness of hospital staff, pain management, and clarity of information given at discharge. The HCAHPS survey typifies this focus by asking patients about their direct experiences with key aspects of hospital care common to a wide variety of hospital admissions that are observable, measurable and actionable.
Positive patient experience can lead to higher satisfaction but the reverse is not necessarily true, as high satisfaction does not guarantee a positive experience with all aspects of care. For instance, a patient may report satisfaction due to a positive outcome or empathetic interactions with a nurse, despite experiencing issues like long wait times or poor communication from other staff members.
Focusing on patient experience is worthwhile because improving it is linked to better clinical outcomes, increased adherence to recommended treatments, and reduced healthcare service use. It is therefore seen as an important indicator of care quality.
Image: Google Gemini
2. “Medical scheme members who score initial hospital stay poorly, more likely to have unplanned readmission” is the headline – but numerous factors affect this relationship.
The Discovery study connects low HCAHPS scores and the likelihood of unplanned readmissions, finding that scores on the question asking patients to rate their stay in hospital (from worst hospital (score 0) to ‘best hospital’ (score 10)) were lower in patients with unplanned admissions. However, causation is not established. Sicker patients might rate their hospital experience worse and be more at risk of a readmission.
Factors known to affect the rating that can be used to make fair comparisons (risk-adjustment) between hospitals when evaluating HCAHPS scores include age, sex, service line (obstetrical, medical, surgical), admission type (overnight vs day admission), emergency status, highest education level, home language and self-rated health.
But surprisingly, when Discovery risk adjusted their results, they found patients with unplanned admissions had higher than expected scores on their overall assessment of their experience in hospital.
“Figure 2 shows that, for those who had an unplanned readmission, actual hospital ratings in 2020 (140.9%), 2021 (103.4%) and 2022 (137.7%) were higher than we anticipated in line with patients’ individual circumstances (especially in 2020 and 2022).”
The factors used in the adjustment may not sufficiently account for severity of illness or co-morbidities, which could be confounding factors affecting both patient experience and the likelihood of readmission.
Mix adjustment and stratification can be applied not only to overall ratings but to all questions in the survey to achieve better understanding of the link between patient experience and readmission. It’s not clear to what extent Discovery have done this.
The Discovery survey is sent to patients 7 days after discharge but the median response time across this sample was 13 days, nearly a week later. This is relevant because patients’ recollections of their hospital stay don’t remain unchanged after discharge, especially for those with subsequent health complications or those who are readmitted. Discovery looked closely at the differences in results of responses at varying times after the index admission. Because response times significantly affect what patients say about their experience they are another factor used to mathematically adjust (“mix-adjust”) publicly reported HCAHPS survey results in the US.
Image: DALL-e
3. Survey response rates are low
Discovery reports a survey response rate of only 11.7% for patients without unplanned admissions, and even lower (6.4%) for patients who were readmitted. These low rates could introduce significant response bias. Reasons for lower response rates among readmitted patients were not explored in depth, and could affect generalisability of results.
4. Overall scores have improved over time – but it’s not clear how
Last but not least, the study shows an encouraging increase in “top-box”[1] scores over the 8 year period, even among patients experiencing a readmission, including during the years of the Covid-19 pandemic, attributing this to improvements in care. A concurrent analysis of changes in hospital policies and procedures, patient expectations, and survey administration methods over time would be helpful to explore and explain this important finding.
“Figure 1: Proportion of patients who gave a top box rating for their overall satisfaction score split between those patients who experienced an unplanned readmission, and those who didn’t, from 2015 to 2022.”
Discussion
Discovery made an important contribution by introducing these validated surveys to SA, by publishing the results and using them to identify possibilities for quality improvement. To support its relevance, they refer to research (from Canada) which finds that a quarter of unplanned readmissions may be preventable.
The study points to the relationship between patient experience scores and unplanned hospital readmissions. Communication gaps in hospital are a potential cause for readmissions, based on bigger scoring gaps between patients with and without unplanned admissions on questions related to communication by nurses and doctors. However, clinical factors and post-discharge care and support might be much more important.
Future research could address limitations by incorporating more detailed clinical data, improving low survey response rates, and using a longitudinal study design to better understand the links between patient experience, quality of care, and readmissions. This may require a closer working relationship between funders, hospital managers and clinicians.
Efforts to improve healthcare quality should focus on all aspects of patient experience as a means to increase overall satisfaction. The ultimate goal should be to provide affordable, safe, effective, timely, patient-centred care, with few unplanned hospital admissions and high satisfaction scores linked to excellent care experiences.
[1] Proportion of patients who give the highest possible rating or response on a particular question.