Background
Quality of care in young persons with chronic disease has been receiving increasing attention over the past years [
1,
2]. Chronically ill adolescents and young adults are faced with the challenges of growing up and starting an independent life in the context of an often debilitating disease. At the same time, there is the additional problem of transitioning from paediatric to adult medical care which will, if poorly staged, result in a care giver gap right at the age peak in risk taking behaviour [
3‐
6]. Consequences may be grave. As an example, non-adherence has been shown to result in a severely compromised prognosis in young transplant recipients and diabetics [
7].
More disorders have by now moved into focus of care improvement programmes including gastrointestinal disease [
8‐
10]. The inflammatory bowel diseases (IBD) comprise ulcerative colitis, Crohn’s disease, and indeterminate colitis or IBD unclassified (IBDu). These chronically relapsing diseases may occur at any age, as early as in infancy, but most commonly in adolescence and early adulthood. The bowel is the primary focus of the inflammatory process; however, any other organ may be involved. General symptoms such as weight loss, delayed growth and fatigue are common, as are issues of psychosocial maladaptation. The high frequency and variability of possible complications, the unpredictability of the course of the disease, side effects of medications as well as the often embarrassing nature of symptoms pose particularly high challenges to affected youth, their families and the attending physicians. In this situation, quality of care is essential to ensure the best possible start into adult life, both with respect to physical wellbeing as well as education, job, and family building.
Various checklists, recommendations, guidelines and consensus statements aiming to improve quality of care have been published, both general and disease specific, including several for IBD [
4,
9,
11‐
14]. Still, it seems implementation lags behind considerably. In consequence, there is a lack of valid data [
15‐
17]. This relates to the description of current health care, to patient preferences and needs, as well as to the evaluation of the effect of care improvement interventions [
17].
Evaluation of quality of care is a complex issue. Classically, structural, procedural and outcome related factors should be considered [
18,
19]. All of these are potentially specific to the medical system, the disease at question and the age group considered. Patient satisfaction is one of several important aspects of quality of care, and possibly the one most difficult to assess and interpret due to its highly individual and subjective nature [
20‐
22]. Patient satisfaction is a significant determinant of adherence and may as such be considered both a particularly important outcome measure as well as a relevant prognostic factor in the care of the young adult [
20]. Simplistically, it may be interpreted as the assessment of structural and procedural quality from the patient’s view. There is a strong interdependency between expectations or personal values on the one hand, and perceived quality on the other hand. Also, patient satisfaction has been shown to be directly related to health status [
20,
23]. Thus, measuring satisfaction with care will ideally be combined with other measures of quality of care and patient wellbeing, depending on the research or policy question at hand. It should also allow for inter-individual differences in patient preferences.
A large variety of instruments is available to assess the quality of care from the patient’s view in children or adults with chronic diseases, some generic, some either age or disease specific [
24‐
34]. Doubts remain as to the applicability and sensitivity in the specific context of IBD patients in the transitional age. For example, for the well-validated IBD specific QUOTE IBD, item generation was based on Dutch persons with a median age of 45 years [
25,
35]. In contrast, the widely used generic CHC-SUN, also available in German, is targeted at younger children [
36]. Two methodologically particularly diligent and informative studies on patient preferences in the US and the Netherlands both focused on adolescents, but only those up to the age of 18 or 19 [
6,
37]. These examples illustrate a common problem in both research and patient care. By either looking primarily at children or primarily at adults youth and young adults tend to be marginalized and little is known about their specific needs and preferences.
For a comprehensive survey on the quality of care in the transitional group of IBD patients aged 15 to 24 years, we developed and tested a specific instrument to measure patient satisfaction. Patient satisfaction was conceptualized as the degree to which aspects of care which are individually perceived as important are met [
20]. The instrument is meant to be used within a multi-modular questionnaire including various indicators of quality of care, sociodemographic variables and assessment of health status.
Discussion
We developed an easy to use instrument with good construct validity and satisfactory test-retest reliability to assess patient satisfaction with quality of care in young persons with IBD. Reviewers have previously highlighted both the high inter-individual variation as well as the “dynamic two-way nature” of patient satisfaction as the “difference between expectations and perceptions” [
20]. This concept is mirrored by our two step approach, as we used the importance patients attach to certain aspects of quality of care to weigh how they rated their experience of care. The structural analogy to similar instruments established in the assessment of adult IBD populations is expected to facilitate acceptance by patient representatives and professionals actively engaged in the improvement of care in these diseases [
25,
35,
47].
The pilot study (sensibility survey) showed good acceptance of the instrument.
In accordance with findings from other studies on patient perceived quality of care [
26], physician competence and empathy were found to be particularly important and are represented accordingly in the instrument. There is some indication that patient satisfaction has marked disease and, possibly, system specific aspects. For example, other than reported from the US or Great Britain [
48] it was evident that waiting times and costs do not play a major role in young patient concerns in Germany and Austria. This is in accordance with informal information from providers in Germany and Austria, and has also been reported from the Netherlands [
4].
A major strength of this study is the rigorous inclusion of the young chronic patient perspective. Patient participation in the development of instruments on patient related outcomes is now generally recommended [
20,
49]. To get an as wide as possible spectrum of different patients, we used a very broad access during the item generating and reduction phase, including multiple recruitment strategies. In particular, by using approaches via the internet and keeping the survey completely anonymous we expected to be able to include youth who might otherwise not have participated. The number and variability of responses by far exceeded our expectation. Also, we were impressed by the very constructive and helpful attitude shown by all respondents, resulting in a wealth of different statements.
We placed a special focus on the quality of the disease specific physician care, reflected by a higher agreement with the global question on satisfaction with the physician in care of the IBD, as compared to the global question on IBD treatment in general. This is in contrast to other instruments, such as the generic Child Health Care Questionnaire on Satisfaction, Utilization and Needs (CHC-SUN) targeted at younger children, which applies a more general approach to medical care [
36].
The instrument facilitates to pinpoint specific areas for improvement based on single item description which may be given as a feed back to the care giver. For example, in our on-going study on the quality of care in the transitional stage, all outpatient departments contributing 20 or more patients will receive the item based correlation matrix as well as benchmarked single item performance. Of note, validation has only been performed for the summary score. In consequence, single item based presentations should be reserved for individual descriptive purposes as suggested.
Limitations
Satisfaction with care is a difficult concept, and some have doubted whether it can be validly assessed at all [
20]. We appreciate that the approach we chose is just one way of tackling this issue. Results may not be used as a direct measure of quality of care, but should be viewed as one of several aspects. Combination with complementary measures of care including those of structural and procedural quality, as well as assessment of health status is recommended to aid interpretation.
Our instrument underwent detailed tests for validation in independent data sets, following well established recommendations [
39,
43,
49]. In contrast, the procedures we used for instrument development deviate in several aspects from standard methods commonly reported. For example, items were generated from open, qualitative internet surveys rather than resorting to, for example, personal expert interviews or patient focus groups [
50]. A broad approach as outlined above was necessary in order to reach a sufficiently diverse patient group which would be truly independent from the subsequent test sample and study population. Unfortunately, so far, there is little empirical evidence on the pros and cons of internet surveys in this context [
20]. Similarly, we used content oriented quota sampling rather than data driven methods to reduce the items. This was, besides being owed to personal preference [
51], in part a consequence of the amount of statements received and the format in which they were available. For example, assumptions were not met to perform factorial analysis. We wish to stress however, that so far there is no evidence that any specific method would have been superior [
52], nor are there standard rules on how item reduction should be performed [
20,
49]. As there was no formal derivation or testing of specific domains, the use of sub-scores is discouraged. Overall, the test performance as reported above proved our approach to be successful, with some minor restrictions:
Comprehensiveness of the instrument was critically assessed by some IBD test persons during the pilot study. We are afraid that this is inherent to the topic, as shown by the large amount of different issues raised by the respondents to the 1st internet survey. Any selection of items will have to be a compromise. Our instrument allows for the appreciation of inter-individual variation by employing weights based on perceived importance of the different items, as used by similar measures [
25,
41]. Still, we suggest, that any study on the satisfaction with care should encourage additional free text comments to accommodate the individual nature of patient preferences.
Another problem relates to a marked ceiling effect. Some of this may be due to our mode of sampling which is likely to have aggravated responder bias. In order to achieve high and timely cooperation rates, we used the first respondents of a larger survey for this validation study. It has been shown that in general, both the particularly satisfied as well as the particularly dissatisfied tend to be more responsive in studies on patient satisfaction [
20]. However, in our case we may have oversampled persons with high satisfaction. This was also observed for other instruments, such as the child ZAP [
30]. Of note, for the QUOTE IBD, a 90% satisfaction rate was assumed normal [
25,
35]. We will rerun the analyses on the distribution of items when using the instrument in other contexts.
While the representativeness of the results is usually a minor concern in studies on instrument development and validation [
39], the low numbers of less satisfied patients posed specific problems with insufficient power when examining agreement of categorized variables and sensitivity to change. There is some indication that stability of patient satisfaction is relatively high and may be more influenced by disease activity rather than actual changes, but more information on this is certainly warranted.
Other practical implications
Our instrument was specifically designed to assess patient satisfaction with IBD physician care. Patient satisfaction, however, is only one element, or outcome, of successful transition, and the physician is only one player determining good quality care. Evolvement of knowledge, self-management skills and maturation in attitudes and behaviour are all necessary to enable the patient to successfully master the challenges of adult life with chronic disease. Owing to the high importance currently attached to transitional care, instruments and surveys are increasingly available for these many different purposes, including the description of the transition process as such [
29,
48,
52‐
54]. We acknowledge that for some aspects qualitative approaches may be particularly helpful [
20,
48]. Naturally, the choice of the instrument will depend on the specific study question at hand.
Conclusion
This newly developed instrument is expected to be a useful tool in the assessment of quality of care from the patient’s perspective in young persons with IBD. It is currently used in two on-going studies on patient care in the transitional stage. More information will be collected on the performance of the instrument in these more heterogeneous patient samples.
Note
This manuscript is based in part on the theses of Andrea Sadlo (Diploma in Psychology, LMU Munich) and Julia Altevers (BA Public Health, University of Bremen). Preliminary results were presented at Annual Meetings of the German Epidemiological Association (DGEpi) (2011, instrument development, 2012, validation) and various working group sessions (DACED 2011, CEDATA-GPGE 2012).
Acknowledgments
The survey was performed in the context of a study funded by the German Crohn’s and Colitis Foundation, DCCV e.V. Representatives of the GPGE and DCCV contributed content expertise to the study and helped with patient recruitment (see also information on author contributions).
O. Freinberger (Ulm/Munich) for setting up the internet survey. We also wish to thank the Rechenzentrum at the Helmholtz Zentrum Munich for internet, J. Heinrich and W. Lohr for general support, and P. Rzehak for sharing of ideas during the planning stages. Members of the German Study Group for IBD (DACED, contact person B. Siegmund), the Association of Gastroenterologists in private practice (BNG, contact person B. Bokemeyer) and the German Language Paediatric IBD registry (CEDATA GPGE, contact person M. Claßen) for participation in the doctor’s survey and recruitment of patients for the internet surveys.
The patient self-help organizations in Germany (DCCV e.V., contact person C. Witte) and Austria (OEMCCV e.V., contact persons E. Groß, A. Hauer) for support and recruitment of patients for the internet surveys, as well as for valuable input in the planning of the study and the drafting of the manuscript.
U. Strauch (Regensburg) and T. Ochsenkühn (Munich) for recruitment of patients from gastroenterology clinics for the pilot study. I. Pelz, A. Bergmann, H. Bruns, E. Bonus, A. Reineke and F. Wehovsky for cooperation in the validation survey, data editing and data entry. H. Pohlabeln for statistical support (BIPS Bremen). H. Zeeb (BA Public Health Bremen) and R. Rosner (Psychology, LMU) for thesis evaluation and feedback. M. Bauer for the translation of the instrument into English. H. Gerds for professional language editing.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
AS planned, set up and analysed the 2nd internet survey, performed the item reduction, set up the final instrument, performed and analysed the pilot study, drafted the manuscript and worked on the reviewer suggestions. JA planned and performed the validation analyses and drafted the manuscript. She also participated in data editing and data entry. JP coordinated the validation survey, data entry and editing. BK consulted on the planning of all surveys, and the interpretation, and facilitated patient access. AB analysed the 1st internet survey and participated in the item reduction. MC coordinated the input of the German Language Paediatric IBD registry (CEDATA GPGE), consulted on the paediatric content and recruited patients during all stages of the project. SK served as the paediatric gastroenterology expert, participated in the expert survey and recruited patients during all stages of the project. AT conceived and coordinated all stages of the project, supervised the theses, and wrote and revised the manuscript. All authors read and contributed to the manuscript and approved of the final version.