International Journal of Quality in Health Care, 2016, 28(3), 275–280

doi: 10.1093/intqhc/mzw024
Advance Access Publication Date: 18 March 2016
Methodology Paper

Methodology Article

The patient satisfaction questionnaire of

EUprimecare project: measurement properties
MARTA CIMAS1, ALBA AYALA1, SONIA GARCÍA-PÉREZ2,3,
ANTONIO SARRIA-SANTAMERA2,3,4, and MARIA JOÃO FORJAZ1,3
1
National School of Health, Institute of Health Carlos III, Madrid, Spain, 2Agency for Health Technology Assessment,
Institute of Health Carlos III, Madrid, Spain, 3Red de Investigación en Servicios, Red de Servicios de Salud Orientados
a Enfermedades Crónicas (REDISECC), Madrid, Spain, and 4Faculty of Medicine, University of Alcalá, Alcalá de
Henares, Spain
Address reprint requests to: Marta Cimas, Escuela Nacional de Sanidad, Instituto de Salud Carlos III, C/Monforte de Lemos
5, 28029 Madrid, Spain. Tel: +34 918222342; E-mail: mcimas@isciii.es
Accepted 2 February 2016

Abstract
Objective: The measurement of patient satisfaction is considered an essential outcome indicator to
evaluate health care quality. Patient satisfaction is considered a multi-dimensional construct, which
would include a variety of domains. Although a large number of studies have proposed scales to
measure patient satisfaction, there is a lack of psychometric information on them. This study aims
to describe the psychometric properties of the Primary Care Satisfaction Scale (PCSS) of the
EUprimecare project.
Design: A cross-sectional survey of patient satisfaction with primary care was carried out by
telephone interview.
Setting: Primary care services of Estonia, Finland, Germany, Hungary, Lithuania, Italy and Spain.
Participants: A total of 3020 adult patients aged 18–65 years old attending primary care services.
Method: Classic psychometric properties were analysed and Rasch analysis was used to assess the
following measurement properties: fit to the Rasch model; uni-dimensionality; reliability; differential
item functioning (DIF) by gender, age, civil status, area of residency and country; local independency;
adequacy of response scale; and scale targeting.
Results: To achieve good fit to the Rasch model, the original response scales of three items (1, 2 and 6)
were rescored and Item 3 (waiting time in the room) was removed. The scale was uni-dimensional and
Person Separation Index was 0.79, indicating a good reliability. All items were free from bias. PCSS
linear measure displayed satisfactory convergent validity with overall satisfaction with primary care.
Conclusions: PCSS, as a reliable and valid scale, could be used to measure patient satisfaction in
primary care in Europe.

Key words: patient satisfaction; primary care, Rasch analysis, scale validation

Introduction central position in medical decision-making [2]. As patients are con-

The measurement of patient satisfaction is considered an essential out- sidered consumers of health services, assessment of healthcare quality
come indicator to evaluate healthcare quality [1]. The change to a through patient’s perspective has become a priority in medical man-
biopsychosocial model of healthcare and the patient-centred care agement. However, patient satisfaction is a complex construct, and
movement has greatly contributed to place patient’s opinion in a despite the rise of several theories in the 80s, there is still a lack of a

© The Author 2016. Published by Oxford University Press in association with the International Society for Quality in Health Care; all rights reserved 275

Downloaded from https://academic.oup.com/intqhc/article-abstract/28/3/275/2237798

by guest
on 11 June 2018
 276
solid conceptual theoretical basis and this hinders its measurement [3].
Linder-Perlz [4] defined patient satisfaction as a positive individual at-
titude to the different dimensions of healthcare services. This attitude
can be influenced by several factors, which can vary depending on the
psychosocial and cultural context. Concerning psychosocial factors,
the discrepancy theory argues that patient expectations are key
to understand the reasons for expressed dissatisfaction [3, 5, 6].
Although there is a controversy about measuring patient experience
of quality of care or patient satisfaction, many authors hold that sat-
isfaction is linked to patient expectations of ideal care and their real
experience of care [7]. Patient satisfaction is considered by most a
multi-dimensional construct that would include a variety of domains:
accessibility, organizational characteristics of the system, physician’s
knowledge, clinical and communication skills or doctor–patient rela-
tionship among others [8].
Questionnaires are the most widely used instruments to assess pa-
tient satisfaction. Although a large number of studies have proposed
scales to measure patient satisfaction in different settings including pri-
mary care and hospitals [9–11], often lacking of psychometric infor-
mation concerning those measures. In a literature review, Sitzia [12]
found that less than a half of the studies reported some validity or re-
liability data on instruments used. In a recent review, Evans assessed
several validated instruments for measuring patient satisfaction and
concluded that more science about validity and reliability of these in-
struments is needed [13]. This information is essential to check that
questionnaires measure the construct for which they were designed.
Robust psychometric data must also be available to properly assess
the validity of the results. In sum, well-validated instruments to meas-
ure patient’s satisfaction are essential to assess this relevant construct.
International comparisons represent a further challenge as research-
ers often translate and adapt questionnaires that were developed in a
single cultural setting to another. Nonetheless, these questionnaires
might not be culturally sensitive. Therefore, efforts to develop a vali-
dated scale that can be used in different countries and translated into
other languages without need for adaptation are especially needed.
Providing quality primary care, the entry door to the healthcare
system in many European countries, has a significant impact on popu-
lation’s health [14]. In the European context, instruments available to
measure patient satisfaction in primary care were either too long, or
developed in a single cultural context. Therefore, the EUprimecare
project developed a short patient satisfaction questionnaire to be
used in the different countries participating in the consortium to be in-
tegrated in a population survey. This project was developed under the
European Commission 7th Framework Program with the objective of
proposing methods to analyse costs and quality of primary care in Eur-
ope [15]. Therefore, the goal of this current study is to analyse the psy-
chometric properties of the Primary Care Satisfaction Scale (PCSS)
developed for the EUprimecare project, using an approach from clas-
sic test theory and Rasch analysis.
Methods
Participants
This study used a sample made up of 3020 patients who responded a
telephone survey. The sample design was stratified polietapic with pro-
portional affixation. Participating countries were Estonia, Finland,
Germany, Hungary, Italy, Lithuania and Spain. Random digit dialing
in strata ensured a randomized sample from the seven countries’ popu-
lations. The study included 431–432 patients per country aged be-
tween 25 and 75 years who had visited a general practitioner (GP)
during the past 12 months due to personal illness or health problem.
Downloaded from https://academic.oup.com/intqhc/article-abstract/28/3/275/2237798
by guest
on 11 June 2018
Cimas et al.
Instrument
The PCSS was developed based on qualitative data, expert opinion
and literature review carried out by researchers from the EUprimecare
consortium. Focus groups were conducted in each country with the
participation of patients and healthcare professionals to understand
the perspective of what defines patient satisfaction with primary care
[16]. Expert opinion was used to select the final list of items. Scores
from 1 to 5 were given depending on the considered importance (1 =
lowest, 5 = highest). The scale was included in a larger population sur-
vey that collected participants’ socio-demographic, health status and
health services utilization data. PCSS comprises 10 items measuring
satisfaction with the following aspects: availability of appointments,
waiting time for an appointment, waiting time in the waiting room,
length of consultations, ease of talking with the GP, listening skills
of the GP, explanation of tests and treatments, preventive activities
and diagnostic test offered and helpfulness of primary care staff.
A 5-point Likert-type response scale, from 0 (very dissatisfied) to
5 (totally satisfied) was used. The survey also included an ‘overall sat-
isfaction with primary care’ item with the same response scale. More
details about the development of the EUprimecare project question-
naire are available in Supplementary material, Appendix S2.
Data collection
PCSS was administered through telephone survey to the selected sam-
ple. Each interview was conducted by personnel experienced and
trained in collecting data through computer assisted telephone inter-
view with Computer Assisted Telephone Interview (CATI) and who
could clarify any doubt. Verbal informed consent was obtained
from all participants prior to interview.
Analysis
Data analysis was conducted in three steps using classic psychometric
approach and Rasch analysis. First, we assessed acceptability of data,
scaling assumptions and factor structure. Second, Rasch analysis was
performed; and third, we analysed construct validity and precision
using the linear measure.
Acceptability was assessed on the basis of data quality (>90% of
fully computable data was considered acceptable) [17], score ranges,
closeness of means to median, floor and ceiling effects (maximum ac-
cepted 15%) [18], and skewness of score distribution (−1 to +1) [19].
The correct grouping of the items into the scale was evaluated using
the item-total corrected correlation considering ≥0.40 acceptable
[20]. Principal components analysis was used to explore the structure
of PCSS. The number of components was determined based on eigen-
values >1 and observation of the scree plot [21].
Rasch analysis is based on item response theory [22]. The model
assumes that the person’s level of satisfaction and the level of satisfac-
tion expressed by the item determine the probability of response [23].
We used the partial-credit Rasch model [24] to evaluate PCSS in a sub-
sample of 300, which allows obtaining accurate estimates of item and
person locations regardless of the scale targeting [25]. For a good fit
between the data and the Rasch model, a non-significant χ 2 P-value
with Bonferroni correction for the 10 items was expected. In addition,
individual item fit residual should be within ±2.5 range. Item and per-
son fit statistics should also follow a normal distribution with an ap-
proximately mean ± standard deviation (M ± SD) of 0 ± 1. Disordered
response thresholds indicate that respondents are not able to discrim-
inate between too many response options, or response options have
confusing labels [26]. In case of model misfit, disordered response
 Measurement properties of PCSS • Patient Satisfaction 277

thresholds are examined to collapse categories and individual item fit functioning (DIF) was used to analyse if the scale was biased by sex, age,
residual range is used to eliminate misfit items. civil status, area of residency and country.
Reliability was evaluated by the Person Separation Index (PSI). This Once a good fit to the model was achieved, the final Rasch model
measure is interpreted similarly to Cronbach’s α: value of 0.70 is required was replicated in two other random samples of 300. Then, Rasch ana-
for group and 0.85 for individual comparisons [23]. The local independ- lysis was performed for the total sample and a linear measure was ob-
ence of each item was examined through a residual correlation matrix tained. This allows converting a row sum score (ordinal scale) to a
where correlations should be under 0.30 [26]. Uni-dimensionality of linear measure (interval scale) of patient satisfaction.
the scale is one of the main assumptions of the Rasch model and was as- Convergent validity was explored through the following hypo-
sessed following the procedure proposed by Smith [27]. Differential item thesis: we expected a moderate to high correlation between the linear
measure and the overall satisfaction item (Pearson correlation ≥0.60).
For discriminative validity we used the known-groups method to
Table 1 Descriptive data of the sample (n = 3020)
check the hypothesis that women, old persons, people living in a
M ± SD rural area, having the same GP or married or living in couple could
n (%) have higher levels of satisfaction with primary care than their counter-
parts [8, 28]. Student’s t test was applied for sex, area of residency and
Sociodemographic characteristics
having the same GP. ANOVA with Bonferroni correction for multiple
Age 50.85 ± 14.13
comparisons was used to analyse marital status and age groups. Pear-
Sex (male) 1502 (49.74)
Area of residency
son r correlation was used to study the hypothesis that patient satisfac-
Rural 1095 (36.26) tion in primary care is increasingly higher for older adults, with higher
Urban 1925 (63.74) number of visits to the GP, more number of chronic diseases or with
Marital status better self-perceived health status [29].
Single 460 (15.32) Precision was evaluated through the standard error of measure-
Married/living in couple 2161 (71.96) ment (SEM = SD × √1 − PSI). The upper limit of the 95% confidence
Widowed/separated 382 (12.72) interval (CI) proposed as threshold to make an approach of the min-
Use of healthcare services in the last year imally detectable change was determined as 1.96 × SEM. PCSS was
Number of visits to the GP 4.28 ± 5.31
considered precise if SEM less than half of SD [18].
Number of visits to the specialist 1.86 ± 3.31
Statistical analysis was performed using Stata version 12.0, SPSS
Having the same GP 2345 (90.86)
Health characteristics
version 19.0 and RUMM2030 for Rasch analysis [30].
Self-rated health status
Very good 349 (11.59)
Results
Good 1343 (44.62)
Regular 1080 (35.88) Main sociodemographic characteristics and information about health
Bad 206 (6.84) and use of healthcare services of the 3020 participants is described in
Very bad 32 (1.06) Table 1. Acceptability parameters and data quality of the measure are
Chronic disease shown in Table 2. PCSS had computable data with items ranging from
Diabetes 276 (9.14) 89.9 to 99.3%. PCSS item scores did not show floor effect, however, a
Hypertension 1054 (34.91)
ceiling effect was present in all items. Skewness was −0.53 and item-
Hypercholesterolemia 760 (25.18)
total correlation coefficients ranged from 0.63 for Item 3 (waiting time
Asthma 180 (5.96)
Chronic bronchitis 252 (8.35) in the waiting room) to 0.76 for Item 7 (explanation of tests and treat-
ments). Principal components analysis showed that PCSS was defined
M, mean; SD, standard deviation; GP, general practitioner. by one component, which explained 50.39% of total variance.

Table 2 Descriptive, data quality and acceptability parameters of PCSS

Item Short description Fully computable (%) Mean Median SD Minimum Maximum Floor effect (%) Ceiling effect (%)

Item 1 Availability of appointments 94.67 4.05 4 0.99 1 5 2.34 39.52

Item 2 Waiting time for an 92.28 3.81 4 1.08 1 5 4.23 30.93
appointment
Item 3 Waiting time in the waiting 99.17 3.69 4 1.08 1 5 4.47 25.28
room
Item 4 Length of consultations 99.17 4.15 4 0.89 1 5 1.04 42.10
Item 5 Ease of talking with your GP 98.44 4.24 4 0.89 1 5 1.31 48.13
Item 6 Listening skills of your GP 99.27 4.27 4 0.86 1 5 0.93 48.70
Item 7 Explanation of tests and 97.65 4.13 4 0.92 1 5 1.59 41.03
treatments
Item 8 Preventive activities 89.93 3.85 4 1.03 1 5 3.39 30.82
Item 9 Diagnostic test offered 89.87 3.96 4 1.00 1 5 2.69 34.82
Item 10 Helpfulness of staff of 91.62 4.26 4 0.83 1 5 0.69 46.11
primary care

PCSS, Primary Care Satisfaction Scale; SD, standard deviation; GP, general practitioner.

Downloaded from https://academic.oup.com/intqhc/article-abstract/28/3/275/2237798

by guest
on 11 June 2018
 278 Cimas et al.

Rasch analysis of the subsample of 300 showed that initial 10-item <0.05. No DIF was observed by sex, age group, civil status, area of
scale had a poor fit to the Rasch model. Items 1 (availability of ap- residence and country. Figure 1 shows the distribution of items and
pointments), 2 (waiting time in the waiting room) and 6 (listening persons in the same logit scale for the subsample of 300. The M ± SD
skills of the GP) presented disordered thresholds indicating that re- person location was 1.699 ± 1.467, indicating that patients showed
spondents were not able to discriminate between response scale op- higher satisfaction level than the average of the scale.
tions 2 and 3; accordingly, both categories were collapsed in one Once fit to the Rasch model was obtained, the analysis was repli-
response option at scoring level. Moreover, Item 3 (waiting time in cated in two other subsamples (n = 300), with similar results. Table 4
the waiting room) showed a poor residual fit (4.247) therefore it shows the transformation of the total raw scores into a linear measure
was eliminated. After applying these changes, a good fit to the for the total sample.
Rasch model was achieved, 9-item PCSS presented a χ 2 (79) = 82.73, Concerning convergent validity, correlation of PCSS with the
P = 0.364 and a PSI of 0.79. Global fit of the initial and final PCSS overall satisfaction item was high (r = 0.70; P < 0.0001). Results of
Rasch models is showed in Table 3. Final model of PCSS is available known-groups validity showed that patient satisfaction in primary
in Supplementary material, Appendix S1. care increased significantly with age groups (ANOVA, P < 0.0001)
Item local independence was observed. Rasch analysis support the and was higher for patients with the same GP (t test, P < 0.0001).
uni-dimensionality of PCSS with a CI lower limit for binomial test Patient satisfaction in primary care had a weak but significant correl-
ation with age (r = 0.20, P < 0.0001). Negligible or no significant dif-
ferences were found by sex, area of residence, number of visits to the
Table 3 Rasch analysis of PCSS: global fit to Rasch model (n = 300) specialists or GP, self-perceived health status and number of chronic
health conditions. Results showed significant differences in patient sat-
Standard Initial model Final model
isfaction by participant country (P < 0.0001). Regarding the precision
Fit residual statistics of PCSS, the SEM was 2.56 (upper limit of 95% CI: 5.01).
Items
Mean 0 0.149 0.156
SD 1 1.876 1.626 Discussion
Persons
Overall, the results of this study showed that PCSS presented satisfac-
Media 0 −0.449 −0.493
SD 1 1.517 1.531 tory measurement properties and represents a valid and reliable meas-
Item-trait interaction ure of patient satisfaction in primary care.
χ2 Low 143.599 82.734 PCSS data quality and acceptability were acceptable, although all
Prob NS 0.002 0.364 items presented ceiling effect. This issue is frequently observed in other
PSI >0.70 0.834 0.796 studies that use patient satisfaction questionnaires [31], although its
Uni-dimensionality: significant t-test reason is still unclear. Many causes are hypothesized, from a lack of
% <5% 5.67 response categories of items to a real high level of patient satisfaction
a
(95% CI) (0.049; 0.048) (0.032; 0.081) of the sample [9]. According to Rasch analysis, items can be ordered
PCSS, Primary Care Satisfaction Scale; SD, standard deviation; Prob,
according to the construct level. Our results suggest that the scale lacks
probability; NS, no significant; S, significant; CI, confidence interval; PSI, items that measure the higher levels of patient satisfaction in our sam-
Person Separation Index. ple. As Sitzia concluded, more research is needed to identify the causes
a
The scale is multi-dimensional if the lower bound of the binomial confidence of patient dissatisfaction instead of working with instruments that
interval is above 0.05. constantly show high levels of patient satisfaction [5].

Figure 1 Rasch analysis of PCSS: person item distribution graph for the final model (n = 300).

Downloaded from https://academic.oup.com/intqhc/article-abstract/28/3/275/2237798

by guest
on 11 June 2018
 Measurement properties of PCSS • Patient Satisfaction 279

Table 4 Rasch analysis of PCSS: raw sumscore to linear measure cause a delay in consultations which often leads to patient dissatisfac-
conversion table tion. The validated EUROPEP questionnaire to assess patient satisfac-
tion in primary care includes an item about time in the waiting room
Raw sumscore Logit scale Linear measure
and in some European comparative studies this item was poorly scored
0 −4.446 0 and presented many differences among countries [33, 34]. This item
1 −3.532 3.298 depends too much on the specific organizational features of the pri-
2 −2.925 5.488 mary care system of each country.
3 −2.521 6.946 The absence of DIF by country suggests that PCSS is free from bias.
4 −2.219 8.035 Therefore, the scale could be applied to measure patient satisfaction in
5 −1.974 8.919
all participating countries with no need for specific socio-cultural
6 −1.765 9.673
adaptations.
7 −1.580 10.341
8 −1.411 10.951
There are currently no instruments considered as gold standard to
9 −1.254 11.517 measure patient satisfaction in primary care and the study did not in-
10 −1.105 12.055 clude another scale for comparison. Nevertheless, we used an overall
11 −0.961 12.574 patient satisfaction measure to determine convergent validity. A previ-
12 −0.821 13.079 ous study had used an item of overall satisfaction with life for the same
13 −0.682 13.581 purpose [35]. PCSS presented a high correlation with the overall pa-
14 −0.542 14.086 tient satisfaction measure. Satisfaction as measured by the PCSS
15 −0.399 14.602 was, as expected, higher in older patients and those having a consistent
16 −0.253 15.129
GP. These results supports construct validity.
17 −0.102 15.674
Our findings are consistent with other studies which have shown
18 0.055 16.240
19 0.221 16.839
age to be a lower but consistent predictor of patient satisfaction [28,
20 0.395 17.467 36, 37]. Older patients generally tend to have lower expectations and
21 0.579 18.131 be more grateful for the attention received than youngers [5]. Having
22 0.772 18.827 the same GP indicates continuity in patient–doctor relationship. Con-
23 0.975 19.560 tinuity in healthcare can contribute to improve confidence and percep-
24 1.188 20.328 tion of quality of healthcare. However, a previous systematic review
25 1.411 21.133 pointed out that relationship between patient satisfaction and continu-
26 1.647 21.984 ity shows much variability [38]. The lack of representation of some
27 1.898 22.890
satisfaction domains in PCSS, for example, treatment efficacy, could
28 2.170 23.871
explain the lack of correlation between patient satisfaction and
29 2.472 24.961
30 2.822 26.224
self-perceived health status, in contrast with the findings of previous
31 3.254 27.783 studies [6].
32 3.857 29.958 SEM has been considered as an estimator of precision, indicating
33 4.700 33.000 the ability of the instrument to detect small differences. In this study,
SEM value shows that PCSS has an acceptable precision.
PCSS, Primary Care Satisfaction Scale. This study has several limitations. First, we could not analyse re-
Note: The raw sumscore to linear measure conversion table provides a way of
sponsiveness due to a lack of longitudinal data. However, precision
obtaining a linear measure from the raw scores. The linear measure is initially
data indicates that PCSS could be a good instrument to detect change.
expressed in logits, and may be rescaled into a more suitable range, in this case
Second, PCSS was included in a larger questionnaire which could have
from 0 to 33. Since Item 3 were removed, and the response category scheme of
Items 1, 2 and 6 was changed, the score range is different than in the initial PCSS produce question order effects. Besides, data were collected by tele-
scale. This table is not valid for cases with missing data. phone interview that lasted over 30 min. Although telephone inter-
view has advantages as low cost and reaching a large number of
geographically spread population, interviews shorter than 20 min
Principal components analysis showed PCSS as a uni-dimensional are desirable to avoid respondent reluctance [39]. Further research
measure of patient satisfaction, which was also supported by Rasch would be desirable applying only PCSS. Third, information about
analysis results. This result is contrary to most of the literature, which time from the last consultation was not available. This variable is im-
presents patient satisfaction as a multi-dimensional construct. Nonethe- portant because level of satisfaction could vary depending on the time
less, Möller-Leimkühler indicated that uni- or multi-dimensionality of lapse from the last consultation to the interview [37]. Fourth, we could
patient satisfaction is not only a matter of statistics results but also of not analyse educational level and income because of differences in cat-
interpretation [32]. Dimensions of patient satisfaction could be greatly egorization between countries. Finally, we considered the controversy
influenced by overall satisfaction, making it hard to determine how of comparing patient satisfaction with primary care between different
many dimensions are involved in. Rasch analysis offers an empiric countries. However, we believe that PCSS could have a potential use
solution to determine the uni-dimensionality of the scale [26]. for European comparisons but further studies are needed.
Rasch analysis results indicate that the 10-item initial PCSS pre- Patient satisfaction is considered also a relevant outcome in clinical
sented poor fit to the Rasch model. Several items with disordered practice due to its relationship with better treatment adherence, which
thresholds were rescored by collapsing response categories and Item improves effectiveness and continuity of care, as well as with better
3 (waiting time in the waiting room) was eliminated. Waiting time health status and quality of life. High levels of satisfaction with health-
in the waiting room is an organizational indicator. In several countries care services have an impact in health for the whole population [31].
such as Spain, most of primary care consultations are overcrowded Therefore, improvement of patient satisfaction should be a priority in
and GP agendas might have a patient cited every 5 min. This could clinical practice and health care policies [6, 40].

Downloaded from https://academic.oup.com/intqhc/article-abstract/28/3/275/2237798

by guest
on 11 June 2018
 280
This study shows that PCSS is a valid and reliable instrument to
measure patient satisfaction in primary care. Research on quality of
health care and patient satisfaction is constantly increasing and
PCSS, as a short, easy to apply and validated scale, could be used as
an outcome measure that provides results in a linear scale.
Supplementary material
Supplementary material is available at International Journal for
Quality in Health Care online.
Acknowledgements
We gratefully acknowledge the EUprimecare project participants, whose contri-
bution made this study possible.
Funding
This work was supported by EUprimecare project that was funded under the
European Commission’s 7th Framework Programme (grant no. 241595).
References
1. Donabedian A. The quality of care: how can it be assessed? J Am Med Assoc
1988;260:1743–8.
2. Laine C, Davidoff F. Patient-centered medicine: a professional evolution.
J Am Med Assoc 1996;275:152–6.
3. Williams B. Patient satisfaction: a valid concept? Soc Sci Med
1994;38:509–16.
4. Linder-Pelz S. Toward a theory of patient satisfaction. Soc Sci Med
1982;16:577–82.
5. Sitzia J, Wood N. Patient satisfaction: a review of issues and concepts. Soc
Sci Med 1997;45:1829–43.
6. Crow R, Gage H, Hampson S et al. The measurement of satisfaction with
health care: implications for practice from a systematic review of the litera-
ture. Health Technol Assess 2002;6:1–244.
7. Keller AC, Bergman MM, Heinzmann C et al. The relationship between
hospital patients’ ratings of quality of care and communication. Int J
Qual Health Care 2014;26:26–33.
8. Serrano-del Rosal R, Loriente-Arín N. La anatomía de la satisfacción del
paciente. Salud Pública México 2008;50:162–72.
9. Mead N, Bower P, Roland M. The General Practice Assessment Question-
naire (GPAQ)—development and psychometric characteristics. BMC Fam
Pract 2008;9:13.
10. Vuković M, Gvozdenović BS, Gajić T et al. Validation of a patient satisfaction
questionnaire in primary health care. Public Health 2012;126:710–8.
11. Granado de la Orden S, Rodríguez Rieiro C, Olmedo Lucerón M del C et al.
Design and validation of a questionnaire of the hospital outpatient clinics in
Madrid, Spain, 2006. Rev Esp Salud Pública 2007;81:637–45.
12. Sitzia J. How valid and reliable are patient satisfaction data? An analysis of
195 studies. Int J Qual Health Care 1999;11:319–28.
13. Evans RG, Edwards A, Evans S et al. Assessing the practising physician
using patient surveys: a systematic review of instruments and feedback
methods. Fam Pract 2007;24:117–27.
14. World Health Organization. La Atención Primaria de Salud, Más Necesaria
Que Nunca. Informe Sobre La Salud En El Mundo 2008. http://www.who.
int/whr/2008/08_report_es.pdf (26 January 2014, date last accessed).
15. European Commission. CORDIS. Final Report—EUPRIMECARE (Quality
and Costs of Primary Care in Europe). http://cordis.europa.eu/publication/
rcn/16625_en.html (19 October 2014, date last accessed).
16. Papp R, Borbas I, Dobos E et al. Perceptions of quality in primary health
care: perspectives of patients and professionals based on focus group discus-
sions. BMC Fam Pract 2014;15:128.
Downloaded from https://academic.oup.com/intqhc/article-abstract/28/3/275/2237798
by guest
on 11 June 2018
Cimas et al.
17. WHOQOL Group. The World Health Organization Quality of Life Assess-
ment (WHOQOL): development and general psychometric properties. Soc
Sci Med 1982;46:1569–85.
18. McHorney CA, Tarlov AR. Individual-patient monitoring in clinical
practice: are available health status surveys adequate? Qual Life Res
1995;4:293–307.
19. Hays RD, Anderson R, Revicki D. Psychometric considerations in
evaluating health-related quality of life measures. Qual Life Res 1993;
2:441–9.
20. Rodriguez-Blazquez C, Frades-Payo B, Forjaz MJ et al. ;Psychometric attri-
butes of the Hospital Anxiety and Depression Scale in Parkinson’s disease.
Mov Disord 2009;24:519–25.
21. Nunnally JC, Bernstein IH, Arellano JAV. Teoría Psicométrica. México:
McGraw-Hill, 1995.
22. Rasch G. Probabilistic Models for Some Intelligence and Attainment Tests.
Chicago: University of Chicago Press, 1960.
23. Tennant A, Conaghan PG. The Rasch measurement model in rheumatol-
ogy: what is it and why use it? When should it be applied, and what
should one look for in a Rasch paper? Arthritis Rheum 2007;57:
1358–62.
24. Masters GN. A rasch model for partial credit scoring. Psychometrika
1982;47:149–74.
25. Linacre JM. Sample size and item calibration stability. Rasch Meas Trans
1994;7:328.
26. Martinez-Martin P, Forjaz MJ. How to evaluate validation data. In:
Parkinson’s Disease Rating Scales. Oxford: Oxford University Press, 2012.
27. Smith EV Jr. Understanding Rasch measurement: detecting and evaluating
the impact of multidimenstionality using item fit statistics and principal
component analysis of residuals. J Appl Meas 2002;3:205–31.
28. Hall JA, Dornan MC. Patient sociodemographic characteristics as predic-
tors of satisfaction with medical care: a meta-analysis. Soc Sci Med
1990;30:811–8.
29. Mira JJ, Aranaz J. Patient satisfaction as an outcome measure in health care.
Med Clín 2000;114(Suppl 3):26–33.
30. Andrich D, Lyne A, Sheridan B et al. RUMM 2030. Perth, Australia:
RUMM Laboratory, 2009.
31. Rebollo P, Cuervo J, Villa G et al. Desarrollo y validación de un cuestionario
genérico para la evaluación de la satisfacción de los pacientes crónicos:
Cuestionario SAT-Q. Aten Primaria 2010;42:612–9.
32. Möller-Leimkühler AM, Dunkel R, Müller P, Pukies G, de Fazio S,
Lehmann E. Is patient satisfaction a unidimensional construct? Factor ana-
lysis of the Munich Patient Satisfaction Scale (MPSS-24). Eur Arch
Psychiatry Clin Neurosci 2002;252:19–23.
33. Petek D, Künzi B, Kersnik J et al. Patients’ evaluations of European
general practice—revisited after 11 years. Int J Qual Health Care
2011;23:621–8.
34. Wensing M, Vedsted P, Kersnik J et al. Patient satisfaction with availability
of general practice: an international comparison. Int J Qual Health Care
2002;14:111–8.
35. Rodriguez-Blazquez C, Frades-Payo B, Forjaz MJ et al. Psychometric prop-
erties of the International Wellbeing Index in community-dwelling older
adults. Int Psychogeriatr 2011;23:161–9.
36. Bleich S. How does satisfaction with the health-care system relate to patient
experience? Bull World Health Organ 2009;87:271–8.
37. Jackson JL, Chamberlin J, Kroenke K. Predictors of patient satisfaction. Soc
Sci Med 2001;52:609–20.
38. Adler R, Vasiliadis A, Bickell N. The relationship between continuity and
patient satisfaction: a systematic review. Fam Pract 2010;27:171–8.
39. Musselwhite K, Cuff L, McGregor L et al. The telephone interview is an ef-
fective method of data collection in clinical nursing research: a discussion
paper. Int J Nurs Stud 2007;44:1064–70.
40. Gupta D, Rodeghier M, Lis CG. Patient satisfaction with service quality in
an oncology setting: implications for prognosis in non-small cell lung can-
cer. Int J Qual Health Care 2013;25:696–703.