The topic is a timely one, as Federal reimbursement will be linked to patient satisfaction scores starting in 2013. Patient satisfaction is an important determinant of future health care utilization, treatment compliance, and willingness to return for care. Patient satisfaction is also a significant contributor to your malpractice risk.
1. Toma G, Triner W, McNutt L. Patient Satisfaction as a Function of Emergency Department Pre-Visit Expectations. Ann Emerg Med. 2009;54:360-367.
This cross-sectional study evaluated consecutive blocks of patients using a pre-visit patient expectation survey, and a post-visit patient satisfaction survey. Their primary goal was to determine if meeting patients’ expectations improves overall patient satisfaction. The surveys were self administered to consecutive patients during periods of block enrollment. ED staff were not aware of the nature of the survey. The initial survey asked about both diagnostic (blood test, Xray, ECG, etc) and therapeutic (pain meds, antibiotics, admission, etc) expectations, and the post-visit survey asked about overall satisfaction with the ED visit as well as other factors associated with satisfaction. Demographic and treatment information was obtained from the patients’ charts.
Of 987 eligible patients, 504 had complete information. Interestingly, 29% of patients reported having no pre-visit expectations. Overall patient satisfaction scores were high, with 50% being “very satisfied and 41% “satisfied”. After adjusting for potential cofounders, there was no association between fulfilling patient expectations and patient satisfaction. Instead, the measured factors with a strong impact on patient satisfaction were the physicians’ interpersonal skills, and to a lesser degree, adequate explanations of the diagnosis, the patients’ perception of ED waiting time, and the satisfaction with the time spent with the physician.
Limitations included the exclusion of very ill patients, some of the wording of the surveys (as Harwood pointed out, “expectations” and “investigations” might not be understood by all patients), Christine mentioned the large number of patients not enrolled, and it was also noted that the surveys were non-validated tools.
Dan G. asked the room about what specific behaviors are interpreted by patients as positive interpersonal skills. Examples from the group: sitting, appropriate touch, introductions/verbal gestures of respect, acknowledging everyone in the treatment room, apologies for long waits, asking if questions or anything else that can be done to make the patient comfortable, frequent updates on status, judicious use of COWS.
2. McCarthy M, Ding R, Zeger S, et al. A Randomized Controlled Trial of the Effect of Service Delivery Information on Patient Satisfaction in an Emergency Department Fast Track. Acad Emerg Med. 2011;18:674-685.
Several past studies have concluded that patient perception of waiting time is more important than actual waiting time. This RCT sought to determine if providing estimates of waiting room times and treatment times would increase overall satisfaction with care. In this study, 1,011 patients triaged to Fast Track (80% of those eligible) during the day on weekdays were randomized to 3 groups: usual care, receiving ED process information, or receiving ED process information + service delivery time estimates (50th and 90th waiting time percentiles). Patients completed a brief survey at discharge describing their satisfaction with care, quality of information received, and timeliness of care. Neither of the interventions (receiving ED process information/service delivery time estimates) affected any satisfaction outcome. Instead, satisfaction was significantly associated with actual waiting room time, and every 10 minute increase in WR time corresponded with an 8% decrease in the odds of reporting high satisfaction with care. Satisfaction ratings also varied significantly among individual triage and fast track nurses and individual fast track doctors.
As Shannon and Matt both commented, this article supports the concept that true patient experience matters more than expectations, although for this study expectations is defined by the information provided to the patients, rather than expectations patients had on arrival to the ED. Matt also stressed the idea that individual medical care providers are all able to impact patient satisfaction scores. Chintan displayed external PressGaney data suggesting that being kept informed is more a predictor than wait times, and both Matt and Christine noted that if the patients had been updated frequently rather than receiving waiting estimates one time at triage, they may have been happier. Harwood described the “Disney Experience”; at Disney wait times are overestimated, and there are frequent markers along the way to let you know how much waiting time is left. Also, using median wait times as an estimate as in this study is not ideal-providing inflated time estimates might have been more effective. Kelly described the signage at Northwestern that provides generous estimates for all ED processes.
3. Hickson G, Clayton E, Entman S,et al. Obstetricians' Prior Malpractice Experience and Patients Satisfaction with Care. JAMA. 1994 Nov 23-30;272(20):1583-7.
Finally, an oldie but a goodie. From 1994, a landmark article-one of the first to test the hypothesis that when it comes to lawsuits, it’s better to be nice than to be good (ok, best to be both).
The authors’ primary objective was to explore the relationship between physicians’ malpractice claims and patient satisfaction. Obstetricians practicing in Florida between 1977 and 1983 were divided into four groups: “No Claims” physicians had no malpractice claims during this time period. “All Others” had a low level of claims, “High Pay” physicians were frequently sued and had large payments against them, and “High Frequency” physicians were sued frequently, but with overall low payments. Birth records from 1987 were reviewed. All cases of fetal/infant death and births with low Apgar scores were included, as well as a number of healthy births. An independent firm conducted interviews using open and closed-ended questions of all available, consenting mothers. A total of 898 interviews were included (63% response rate). Patients of the Hi Frequency physicians were significantly more likely to state that they felt rushed, never received explanations for tests, and were ignored compared to the other physician groups. Problems with physician-patient communication were the most frequently offered complaints. Patients of No Claims physicians were consistently the most satisfied with their care.
Of interest, the No Claims and All Others groups included substantially more perinatal and neonatal deaths and therefore might have led to more negative feelings towards those doctors, yet patients of these physicians were more satisfied with their care. Another hypothesis of why doctors are more frequently sued is that they provide technically inadequate care, however the authors abstracted a subset of medical records and found no significant differences in the technical care provided.
Janna pointed out the limitation that the interviews took place 5 years after the event, which may have biased or lessened the specific memories of the mothers. The study also provided limited specific information about the physicians and their practices. Finally, this is a study of obstetricians, however the results have good face validity and are likely able to be extrapolated to other specialties.