Using patient education to manage expectations may be the only way to address some HCAHPS survey issues because of factors that are built into the questionnaire.  The design process for the HCAHPS survey resulted in some questions that are difficult to completely address with operational improvements.

The survey problems are unavoidable because each hospital must use the standard survey, without any ability to adjust questions to match situations that are unique to them or correct test biases that are negative to them or clarify question intent.  The only tool they have in these cases is management of expectations which can best be done under a consistent and uniform patient education program.

Patient education has been proven to be effective at improving outcomes but an internet search shows that most HCAHPS improvement programs focus on staff and operational issues, with minimal, if any, changes in standard patient education programs.   Efforts to establish realistic patient expectations through a uniform, planned patient education program seem to be lacking and there are no publications that combine “patient education” and HCAHPS in the MEDLINE/PUBMED medical publication search engine. 

Understand HCAHPS Flaws

Understanding the flaws in the HCAHPS survey requires knowledge of both the national scores and the correlation of ratings and recommendations with the individual questions. A study[1] in the New England Journal of Medicine looked at these correlations and this data is given below along with the average national hospital scores for what is called top box scores (definitely, always or 9-10 ratings).

The scores are highly correlated yet the correlation is lower for recommendations than for ratings.  This must be because the HCAHPS survey does not completely cover the reasons people make recommendations.

In addition, the correlation varies significantly between questions, with “Nurses communicated well” which had relatively high performance of 74% “always” and a correlation of .70 with recommendations compared to another important operational question “Patients received help as soon as they wanted” which had had the lowest “always” at 62% and lowest correlation, .46.

Ratings and Recommendations Correlations to Individual Questions
 Compared to National HCAHPS Scores

Manage Specific Issues

There are at least three specific issues that cause these variations, affecting half of the HCAHPS survey (9 of 18 evaluation questions.)   Active management of patient expectations is an important way to correct these issues as detailed below:

1.       Realistic Patient Expectations of Service.  It is almost impossible to get improvement in some areas without realistic patient expectations as to what good performance is.  There are three service questions in the HCAHPS survey that require establishing what patients can expect.

First and most important is call button response, reported on Hospital Compare as “Patients received help as soon as they wanted.”   This question has one of the lowest national 9/10 top box scores (62%) and also has the lowest correlation with patient ratings and recommendations. 

This is an issue of “wants” (what you would like) versus “needs” (what you must have.)   Patients may think that call button response does not meet their “wants” but realize that their desires for rapid service are not very important compared to the “needs” of themselves and others that keep nurses from responding immediately.   Any hospital that tries to delivery cost effective care has to focus on meeting “needs” rather than the 5 star hotel “wants” service that this question asks about.

Simply shortening call button response time may not improve scores significantly because the “wanted” time could be zero or close to it.  Getting a meaningful improvement in this score requires explaining to the patient what reasonable call button response is so that they form a realistic expectation where their “wanted” response time is close to the “needed” response time that nurses are trying to meet.  This will increase patient satisfaction while making the scoring more likely to reflect this expectation and increase its usefulness to both patients and hospitals.  Note that there is a balancing act in explaining why the response time is not immediate as one that is too long will likely lower both this score and recommendations. 

The two other service questions that need expectations established are the two identical questions for Nurses and Doctors on treating patients with “courtesy and respect.”  While most people are experienced in evaluating whether a waiter is treating them with “courtesy and respect”, they are probably not used to evaluating this during the rushed activities of a hospital, particularly when treatment is underway or patients are confused from medication or illness.  The problem is that testing the concepts of courtesy and respect is done with a multipoint questionnaire so a single question is inadequate to identify issues.

This actually affects 6 total questions as “courtesy and respect” scores are blended together to with questions on whether Doctors and Nurses “listen carefully” and ”explain things” to make a composite “communicate well” score that is available to patients.   Nurses’ “communicate well” had the highest correlation with hospital ratings and recommendations of any question, showing the importance of the nurse’s role to patients.

The opportunity for improvement in care delivery is to clarify to both patients and staff how “courtesy and respect” is done at the hospital.  For example, educating and reassuring the patient and family on respect concepts such as “treatment with dignity” will add to their comfort with treatment and increase satisfaction[2].   Finally, linking the concept of courtesy and respect to the concrete actions in the later two questions of “listen carefully” and “explain things” allows patients to evaluate a performance that can be divorced from the busy activities of a major hospital delivering cutting edge, efficient care.

2.       Realistic Patient Expectations of the Hospital Environment.  Modern hospital care ideation in America is usually done on the basis of “absolute best care possible” without much consideration of efficiency.  The questions in the HCAPHS survey reflect this bias.  Except for pain management, questions are worded as absolutes that minimize the likelihood that patients will use their perspective (contrary to the stated intent of the survey.)  

The “quiet” question in the HCAHPS survey is the most affected by this lack of value orientation.  Hospitals are busy, relatively noisy places. The World Health Organization guideline value for background noise in hospital patient rooms is 35 dB, with average noise levels (LAmax) not to exceed 40 dB, about the noise level of rainfall.  The question in the survey is phrased as “how often was the area around your room quiet at night?”  A factual answer to this question might be “Never” even though the patient slept well.  This might be the reason this question has the second lowest correlation to ratings (.56) and recommendations (.48).

If asked on a perceptual basis, such as “How often was the area around your room quiet enough for you to rest well at night” the answer would be more meaningful as there is research that shows better patient outcomes when noise is reduced to a level that does not disturb patient rest.  The opportunity is for hospitals to define “quiet” in the hospital setting so that patients relate it to their rest and well-being, not extraneous noise levels.

This is particularly important to existing hospitals that try to deliver cost effective care.   Studies have shown that there are benefits to the lower noise levels of single rooms but others indicate that there may be offsetting benefits[3] from the socialization in semi-private rooms.  Consequently, the incremental cost of private rooms over semi-private is probably not justified by sound reduction alone if there are not other medical reasons for isolation. 

Most important is that addressing this question should result in improved rest in an unfamiliar environment by educating patients that many of the sounds they hear are due to the efficient delivery of care and are normal in a hospital.  The secondary benefit is focusing them on judging the sounds that disturb them and increasing the meaningfulness of the HCAHPS “quiet” score.

3.       Minimizing Survey Order Bias. The HCAPHS survey is a standard questionnaire where questions are asked in a fixed order.  First are scaled questions (Always, Usually, Sometimes, Never, or Yes/No) then patients are asked for a hospital rating (from 1 to 10) and a recommendation (Definitely No, Probably No, Probably Yes, Definitely Yes.)  This generates what researchers call “order bias” where the questions before influence the later answers. 

The NEJOM article cited above has shown these questions to be highly correlated with one another, an indicator of order bias.  Another indicator of order bias is that hospitals that have strong images and reputations seem to get higher recommendations than their rating scores would merit. This might be because the HCAHPS survey does not completely cover the reasons people make recommendations.  Ratings and recommendations are always asked at the end of the survey so hospitals that don’t have strong reputations would tend to have recommendations that are directly related with ratings.

It would be best if questions were asked in a random fashion to minimize order bias and increase the survey accuracy but this is unlikely to happen.  Informing the patient of the hospital efforts and intent in appropriate areas will minimize order bias caused by lack of knowledge/reputation.  An added benefit is that patients will know more about the positive activities of the hospital which will increase the potential for future recommendations outside the HCAHPS survey.

The HCAHPS survey has serious structural issues that need to be addressed through patient education that establishes expectations for the individual care areas.  Traditionally, this is been left to the individual nurse and doctor to communicate but the opportunity is to establish a consistent patient communication/education program.  A consistent program could have measurable goals and outcomes that would be subject to review and improvement.  The bonus is that a successful program would build the hospital’s reputation and lead to increased patient satisfaction.

 *Sidebar:  What can you say to Patients about HCAHPS?

Before we talk about using Patient Education with the Hospital Consumer Assessment of Healthcare Providers and Systems (HCAHPS), we have to address whether it is even allowed under HCAHPS guidelines. After all, using Patient Education to increase survey scores seems like it might be directly contrary to the intent of the survey. The expectation on what the HCAHPS survey is measuring is clear: 

…to the fullest extent possible…patients’ responses are informed only by the care they received during the hospital stay.

                                                                 HCAHPS Bulletin Number 2009-01 Revised

To make sure that hospitals understand what they can do, the Quality Assurance Guidelines include a section on “Communicating with Patients about the HCAHPS Survey” shown in Table 1.

Further clarification was given in HCAHPS Bulletin Number 2009-01 Revised, The Use of HCAHPS in Conjunction with Other Hospital Inpatient Surveys which was quoted in part above.  The comments that guide how patient expectations can be managed are:

“In general, activities and encounters that are intended to provide or assess clinical care or promote patient/family well-being are permissible.

However, activities and encounters that are primarily intended to influence how patients, or which patients, respond to HCAHPS survey items should be avoided.”

A strict interpretation of HCAHPS guidelines would be that patient education (that is, direct patient communication) can be used to improve HCAHPS survey scores as long as they avoid using any of the HCAHPS specific terminology and are intended to primarily improve patient outcomes rather than increase survey response rates or scores.