Why is the rating for overall patient satisfaction much
lower than ratings for most, if not all, of the other survey
ratings questions?

Response:

It is quite possible for this to occur from time-to-time, as well as the opposite; a high rating for overall satisfaction and relatively low scores for most other items, such as measures for nursing care, service waits, communication with staff, etc.  HealthStream Research asks overall patient satisfaction as a separate question.  It is not simply a roll-up of all other survey items.  We believe that it is important to obtain a measure of satisfaction that is truly “overall”, as opposed to a roll-up or the “grand mean” of other questions. 

This methodology makes comparisons to our National Database stronger than it would be if we simply did a rollup of all other questions.  For example, HealthStream Research clients are allowed flexibility in customizing their surveys to meet their needs.  Because of this, overall satisfaction for various clients would be based on different measures if we simply did a roll-up of all questions asked, making comparisons to a “standard” quite difficult. Another problem with rolling up questions to determine overall satisfaction is that when this is done, you are giving the same weight to questions that really don’t drive patient loyalty as to those that do.  Thus, you could score high in this rollup because you score exceptionally high in non-predictive questions (such as food quality), but still have high degrees of patient dissatisfaction because you are not scoring as well on the issues more predictive of patient satisfaction.  Using one overall satisfaction question prevents this.

HealthStream Research sees the occurrence of a low overall satisfaction rating relative to scores for other items from time to time, particularly at the unit level where sample sizes are not as large.  It is rare to see this at the “hospital” level and it rarely occurs over a long period of time.  Hence, a fluctuation in survey results due to lower sample sizes is a primary cause.  Using annual data at the unit level typically reduces or eliminates this issue.

There are other reasons that account for what on the surface appears to be inconsistency in scores.  For example, it may be that all we are measuring might be rated highly, but some other factor or factors that are not being measured cause one or more patient to be dissatisfied.   This situation would be reflected in the overall satisfaction score but not the other scores.   

The opposite can also occur.  A unit may perform average to poorly in many facets of patient care service, but something could have happened (an employee who did something extraordinary for example) to cause the patient to be "completely satisfied".  We can't account for all situations that could cause a patient to be highly satisfied or highly dissatisfied.  These types of situations can happen and the correlations to overall satisfaction can still be high overall.  As indicated, these situations tend to even out on an extended period of time, but can be frustrating on a quarterly basis.

We cannot measure the hundreds, if not thousands, of factors that could ultimately influence overall patient satisfaction.  Based on years of experience, we recommend questions that we know are the most predictive of patient satisfaction.  But these questions will be more or less predictive on a hospital-by-hospital basis as all organizations are unique.  When these occurrences happen, there are a number or routes to take, including:

• Add a follow-up question to determine issues not in the survey that may be influencing overall patient satisfaction
• Substitute questions with other known “key” drivers of overall satisfaction
• Hold focus groups to determine issues not identified in the survey 

The medical outcome of the patient visit is another factor that can influence overall satisfaction, but has very little impact on ratings for less global items like, “nurse kindness”, “the registration process”, “waits for services”, etc.  The unit’s staff may have done an excellent job in measures that relate to customer service, but a less than desirable medical outcome would detract from the great service provided and be reflected in the overall satisfaction score.