Dear Colleague

Thank you for your interest

If possible, a questionnaire to see patient satisfaction for health services provided to them as they slept at the hospital.

I think the following study, based on the Competing values framework offers an interesting and theoretically grounded approach: Ancarani, Di Mauro & Giammanco (2009). How are organisational climate models and patient satisfaction related? A competing value framework approach. Social Science & Medicine, 69, 1813-1818.

We use a patient satisfaction survey based on Picker institute

http://www.pickereurope.org/

Thank you all. As a matter of fact I started using the Picker Patient Experience 15 (PPE-15) questionnaire.  It fullfils my requirements:

So Martina and Marianne, yes the PPE-15 happens to have been my best option. I had already collected some interesting data from a local community hospital in Mexico. Thanks for your comments.

Dear Ministry of Health in Palestine. I am unable to spell your name with our characters, but please consider using the Picker Patient Experience 15 (PPE-15) questionnaire which is fully reported in The Picker Patient Experience Questionnaire: development and validation using data from in-patient surveys in five countries

CRISPIN JENKINSONa, ANGELA COULTER1 AND STEPHEN BRUSTER1

which was publsihed in International Journal for Quality in Health Care 2002; volume 14. Number 5: 353-358

I think you are going to find it  useful for your purposes. Also contact the authors and the Picker Institute to receive further help.

Good luck!

The U.S. agency AHRQ developed HCAHPS. It is a blend of experience of care, satisfaction, and quality of care indicators. The instrument is in the public domain and it has been in use long enough to have comparative information also available. 

Hi, I sincerely apologise if I am speaking about issues that you are already aware of.

Is there a reason why you are restricting this to quantitative methods?

Quality is a multidimensional and highly context-dependent concept. However attractive (in terms of e.g. cost and convenience), I believe that a one-off, fixed perspective snapshot of quality is insufficient. Quality measurement and continuous quality improvement are interdependent. 

If soliciting the views of potentially vulnerable people, this should be done with a good evidence base for the appropriate tools and methodology. All of this is context dependent. Even if using a validated quality of care questionnaire, I would suggest that this is not enough on its own.

I would certainly steer away from "satisfaction" surveys. Given the potential diversity of the survey group, a simple satisfaction survey is not likely to provide any useful information. Here are some reasons why:

1) In terms of effectiveness, read, e.g. "Ineffective -- That's the Problem with Customer Satisfaction Surveys" (http://asq.org/qic/display-item/?item=12632)

2) The notion of "satisfaction" occupies a tiny portion of the continuum from "awful" to "absolutely amazing", so is of limited use. See Fred Lee's papers, e.g. "Stop Measuring Patient Satisfaction" (http://www.ncbi.nlm.nih.gov/pubmed/12800626).

3) Satisfaction is usually rated high in surveys (e.g. see this paper: "A Demonstration of the Impact of Response Bias on the Results of Patient Satisfaction Surveys" http://onlinelibrary.wiley.com/doi/10.1111/1475-6773.11194/pdf). As this paper points out, this is worsened by the effect of "satisfaction inflation" being greater for those who receive a poorer service.

4) From a methodological perspective, getting the "wrong answer" from a survey is easy.  As I'm sure you are aware, surveys involve sampling and as such are affected three phenomena; chance, bias and confounding. Chance. If you are lucky you might have a big sample to reduce the effects of chance by increasing statistical significance (but "statistically significant" and "meaningful" are not the same things). Bias can be introduced in many ways. To name a few; sampling bias (you asked an unrepresentative group and generalised the answer), response bias (you asked a representative group but an unrepresentative sub-group responded), bias in interpretation (sub-groups interpret the question in different ways), and so on. There are so many - all of which should be explored and their effects incorporated in both the sampling approach (to reduce the effect) and in the interpretation of results (to understand and incorporate the impact). Confounding. Here, you are actually (and unknowingly) measuring a different causal mechanism.

As you can guess I am not a fan of blunt satisfaction surveys - there is sufficient research that supports the avoidance of this approach for anything but the simplest scenario.

It might be worth looking at more sophisticated tools, like Kano's very powerful "paired questions" approach (which has its roots in marketing). I would also look into touch point analysis to identify the key service interaction points and ask specifically about them (from an improvement perspective). In commercial marketing, instances where the customer has contact with product or service “system elements” (whether direct or indirect) during their experience are referred to as “touch points” (see the work of Meyer & Schwager, 2007 and Brugnoli, 2009 or visit http://www.servicedesigntools.org/tools/108). Data collected about the customer’s experience at these instances can inform “touch point analysis”. A series of touch points represents the customer’s pathway of experiences with the product or service. Beyond commerce, Golaszewski et al (2008) describe touch points as those formal and informal processes that reinforce culture (I say this as you state an interest in organisational culture).

I don't believe that it is unusual for surveys to be deployed without any recognition of the issues I've highlighted (and there are more!) - and then for key strategic/operational decisions to be based on the findings.

I'll stop bleating now! ;o)