The Emergency Department Patient Experience: In Their Own Words.

In this paper, the authors review the verbatim comments from patients' post-emergency department discharge survey results and highlight the impact that doctors and nurses have on the patient's experience. This paper also explores the benefits of delivering a positive experience on patients, clinicians, and healthcare systems. Many interventions that have worked are simple to implement and do not tend to require significant capital investment while having the potential of improving emergency department encounters for all those involved. There are always limitations, selection biases, and issues with generalizability in reviewing retrospective and subjective survey response data.

Introduction

Emergency Department (ED) patient experience has become a critical differentiator as departments compete for patient visits and seek to deliver optimum care. The Agency for Healthcare Research and Quality defines patient experience as “the range of interactions that patients have with the healthcare system…from doctors, nurses, and staff in hospitals, physician practices, and other healthcare facilities.” (1) Satisfaction with the patient experience is the net of patient perception and expectation, that is, a patient will be satisfied with their visit if their perception of the experience is greater than their expectation for that visit. To improve patient experience, one must either enhance the perception of the experience, manage patient expectations, or ideally do both. EDs are uniquely positioned in their ability to shape a patient's first impression as they navigate the healthcare system. However, the challenges in delivering a positive experience cannot be understated. Patients will often describe the day they interact with an ED as the worst day of their lives. This preconceived subtext makes delivering a positive experience that much more difficult.

Previous literature has shown that clinical outcomes are often correlated with patient experience (2). However, this has not always proven to be true (3) and understates that patient experience remains a multifactorial problem that requires a multidisciplinary approach. With patient satisfaction tied to reimbursement, finding evidence-based interventions that have previously worked becomes all the more important. In this review, we will discuss the benefits of delivering an exceptional patient experience. These include both benefits to the patient as well as the physician. We will also discuss interventions that have been shown to improve patient experience and common interventions that have not had their effectiveness confirmed.

Clinicians benefit from delivering a positive patient experience (Figure 1).

Figure 1.

Benefits of delivering exceptional patient experience.

A retrospective look at a correlation between patient satisfaction and patient complaints, risk management episodes, and rates of malpractice lawsuits found that decreases in physicians’ patient satisfaction survey scores from the highest to the lowest third were associated with a statistically significant increase in unsolicited complaints from patients and risk management episodes (4). The data, however, did not find a statistically significant difference between the highest and middle third, focusing on the bottom third that sees malpractice lawsuits at a relative risk of 2.10 times that of the highest third.

The healthcare system benefits from providers delivering a positive patient experience. A retrospective cohort study in 2017 found a correlation between a positive patient experience and increased profitability as well as an even stronger correlation between a negative patient experience and decreased profitability (5). In this study, across almost 20,000 observations, only a 1% increase in patients who would definitely recommend the hospital was associated with a $1,072,000 increase in net patient revenue and an almost $250,000 expected increase in net income. The study's recommendation was that these increases in income would justify incurring costs associated with bolstering patient experience programs.

Methods

Our ED uses the Press Ganey patient satisfaction surveys for benchmarking physician performance. These proprietary surveys consist of 36 questions across eight sections: Arrival, Nurses, Doctors, Tests, Family or Friends, Personal/Insurance Information, Personal Issues, and Overall Assessment. The scores can range from 1 (very poor) to five (very good). The percentage of responses answered “very good” is called the Top Box score.

Our ED system comprises two hospital EDs, and three freestanding EDs. One of our hospital EDs is a level 2 trauma center and home to residency training programs in emergency medicine, surgery, internal medicine and transitional year. The other is a very busy community hospital. Together, our five EDs see over 300,000 patients per year. Post discharge surveys are sent by a third party to patients discharged from the ED. All post ED discharge surveys from patients seen between September 1st 2020 and October 31st 2020 were analyzed using the word cloud analysis function in JMP 14.0 for the Macintosh.

Results

A total of 1481 surveys were returned during the study period, of which 48% were from the hospital EDs and 52% were from the freestanding EDs. The survey response rate was approximately 15%, similar to response rates at comparable institutions. There were four possible reactions to the questions, including positive, negative, neural, and mixed. For purposes of the analysis, neutral and mixed were combined. The verbatim word clouds are presented in Figure 2.

Figure 2.

Summary of word cloud analysis demonstrating that “doctor” is an important word in the patient experience and especially for negative and neutral reviews.

A word cloud is a visual representation of word frequency. The more commonly the term appears within the text being analyzed, the larger the word appears in the image generated. What is immediately obvious in all of the reactions is that the largest word is doctor. This suggests that a patient's ED experience is largely based on their perception of their doctor. The word cloud for “negative” was perhaps most striking, with the words: “doctor” and “never” being most prominent. We then reviewed the individual patient comments in their entirety to identify actionable items. Using this information as a springboard, we retrieved the individual physician data and developed a coaching plan targeting the bottom 30% of performers. The coaching plan involved individual observation of physicians, and real-time feedback regarding how their interaction could have been more optimized. Figure 3 is a patient engagement script that was developed for use at our facilities at the height of the COVID-19 pandemic, when the patient experience was sometimes further compromised by the necessary social distancing guidelines.

Figure 3.

Emergency department scripting tool.

Discussion

Overall, the three word clouds demonstrate that clinicians hold considerable influence in affecting a patient's experience illustrated by the size of “doctor” in the word clouds. “Nurse” is almost as large supporting those interactions that staff, clinicians, and nurses, have with patient are remembered and heavily affect the patient's perception of the encounter and their responses to standardized questions. Interventions should seek to empower and improve those interactions between staff and patients.

As part of our intervention to improve Top-Box scores, we incorporated a number of techniques to improve survey results. In the time of COVID-19 and universal masking, framing the frustrations of limited communication and facial expressions around patient and family safety can work to provide prospective of measures taken during a global pandemic and why patient interactions and the care that is delivered has changed significantly. Scripting is a technique to purposefully use the same language during the patient encounter as will be seen on the post-ER visit survey to trigger your efforts at meeting those needs. For example, “I am ordering pain medication for your comfort” directly addressed the survey question, “Doctors concern for your comfort while treating you.” Managing up your team is a technique to improve the perception of the caregivers a patient will interact with. “Amy is an amazing nurse. You are lucky to have her taking care of you today” will instill confidence in both the patient as well as the nurse taking care of the patient. Managing up works best with involvement from everyone involved in the care of the patient: clinicians, nurses, radiology, lab technicians, etc. Keeping patients updated with rounding or white-boards directly addresses survey questions as well as meeting the patient's psychologic need to understand the waiting process and that they are being cared for and not forgotten. Frequent re-evaluations after interventions and sharing testing results allows the patient to take ownership over their care and provides tangible evidence that something was done for them. As we collect post-intervention data, we hope these interventions will lead to an improvement in survey result scores.

Limitations

Word cloud analysis has inherent limitations. For example, rather than viewing the meaning of the whole comment, it focuses specifically on individual words. Word clouds can misestimate issues because similar words are not combined and the patterns in the cloud can be out of context. Importantly, word clouds cannot inform how important the word is relative to a given outcome such as compliance or provider recommendation. Despite these limitations of word cloud analysis, in the current study, the word “doctor” as the most frequent word in all comments does provide insight on the individual that might be the most important focus for behavioral change in the ED context. There is evidence from quantitative analysis of ER survey data that ratings of the physician account for the greatest explained variance of overall service satisfaction (6).

There are a number of limitations in using survey response data to make meaningful progress in understanding the Patient Experience problem. The Press Ganey tool itself is a significant limitation, as its proprietary nature means that the randomization and distribution of the survey and dissemination of results is not transparent. Numerous studies have noted the problems with nonresponse bias associated with patient satisfaction surveys (7–9). Furthermore, the sample size of Press Ganey surveys can create unacceptable margins of error (10). Throughout our analysis we also found a disconnect between responses and survey scores. Even positive responses do not always correlate to top-box scores. Future research will have to specifically survey patients for reasons they did not find their ED experience to be top-box worthy. This can be done through call-back programs which can offer another layer of data as well as an opportunity to improve survey scores which might add an issue of confounding. An additional limitation is the selection bias inherent in survey responses. In an orthopedic setting, research has demonstrated rates as low as 16% with a predominance of responses from advanced age patients (6). An ED study quotes a response rate of 12%, and states that is similar to comparable peer institutions (11). Understanding the demographics of the patients that fill out surveys can be helpful to plan interventions around. That level of granular data was not available for this analysis.