} Stirs Carv ijtutl

Vol. 20, No y pp 2>()->.M)

Koil Williams s. lm:

A Psychometric Analysis of the

Patient Satisfaction With
Nursing Care Quality
Questionnaire
An Actionable Approach to
Measuring Patient Satisfaction
Heather Spence Laschinger, PhD, RN;
Linda McGiUis Hall, PhD, RN; Cheryl Pedersen, MSc;
Joan Almost, MScN, RN
Patient satisfiction with nursing care quality- is an important indicator of the quality of carr pro-
\-idcd in hospitals, lliis study icstcd a newly dcvclupt-d patieni-tcntcrcd measure of palienl satisfac-
tion with nursing care quality within a random sample of !t hospitals iii Ontario, Canada. Results
of this study revealed that the newly divt-lopecl instrument had excellent psychometric proper-
ties. Ibtal scores on satisfaction with nursing cart- were strongly related to overall satisfaction with
the quality of care received during hospiialization. The rt-suks of this study yielded actionable,
patient-focused asults that can be used by managers to addnrss areas requiring inipnn-cmcnt. Key
word5: insininient deit'lopmcnt, nursing care; patient satis/action, qtiuiity of caw

P ATIENT satisfaction with nursing

quality is an important indicator of the
quality of care provided in hospitals. This
care they experience during hospitali/ation
are an important component of assessment
of overall hospital quality. Tliis is supported
is not surprising, given the significant n)le by previous research th:it has shown pa-
of nursing in the patient care experience. tient satisfaction with nursing CAVC to be the
As the recipients of nursing and medical stnmgest predictor of dominant itifluence on
care, patients' perceptions of the quality of patient overall satisfaction with their hospital
experience.'"^
However, tnany patient satisfaction instioi-
ments measure only healthcare providers' per-
Prnm the School ofNurstnu. University of ceptions of what elements of patient care
Ontario, lonilon (i)r l^iscbinger): Ihiitvrsity of
Toronto (l)r lltAil ami Mas l\-dersen and Almost): are valued by patients. Failure to take into
Nursing Health .Se/Tici'S Research I'nit. Toronto (Drs account factot^ identified by patients as im-
iMschin^er and Hail): mui the Canadian Institutes of portant to their care is inconsistent with the
Health Research (Dr Hall). Ontario. Canada.
notion of patient<entered care and patient
ms project wasfunded by tlje Hos{>itul RefHtrt Cotlabo- empowerment. In today s consumer-oriented
raliiv: Ontario Hospital Association ami Onitirio Min-
istry' ofHeiUth and Long-Term Care. healthcare markets, a patient-centered mea-
Com-spofuiinji author: Heather .S/wme lMschinf>er,
sure of satisfaction with the quality of nursing
Phi). RN. Scijoui of Sursinn. l/niivrsity of Western On- care received is an important component of
tario. London, Ontario. Canada N6A 5CJ (e-mait hbl® hospital quality management systems. In ad-
uut>.ca). dition, to be useful tools for quality impnwe-
AcceptedfijrptdJiication: July 13, 2004 ment purposes, patient satisfaction measures
220
Psychometric Analysis of the Patient Satisfaction Questionnaire 221

must also pr<Kluce "actionable" results. The to influence satisfaction ratings. That is, pa-
puq>ose of this study was to test a newly de- tients with good health status postdischarge
veloped patient-centered measure of patient report greater satisfaction than those with
satisfaction with nursitig eare quality. poor health status.'^^•'* Nurses" job satisfac-
tion has also been shown to be related to pa-
tient
CONCEPTUALIZATIONS OF PATIENT
SATISFACTION Consequences of patient satisfaction
with nursing care
Patient satisfaction is often defined in terms
of the match between patients' expectations Patient satisfaction with nursing care has
of care and the acuial care received.^''* Pa- been consistently found to be the strongest
tient satisfaction with nursing eare has been predictor of patient satisfaction with the over-
defined as the patients' subjective evaluation all healthcare experience.'- Several studies
of the cognitive-emotional response resulting have shown that patients who are more sat-
from the interaction between the expecta- isfied with their care are more likely to fol-
tions of nursing care and the perception of low medically prescribed regimens.^"'^^ More
actual nurse behaviors/characteristics.'** A satisfied patients are also more likely to rec-
similar defitiition was proposed in the health ommend the hospital to family and friends^'"
services literature by Cleary and MeNeil." and have greater intentions to return to the
In the health services marketing literature, hospital in the future as opposed to select-
patient satisfaction is viewed as a mediator ing another hospital.*' Finally, Jaipatil and
between patients' perceptions of quality and Rosentha!*- found that hospitals witb higher
their future intentions to reuse the service or patient satisfaction with nursing care liad
recommend the service to others.'^ lower severity-adjusted mortality.
Measurement of patient satisfaction
Related literature with nursing care
Antecedents of patient satisfaction witb A valid and reliable patient satisftaction
nursing care measure used on a regular basis in formal
Many factors have been show^n to have an health services evaluation systems can pro-
impact on patient satisfaction. Personal char- vide meaningful comparable data over time
acteristics of patients, such as cultural back- and form the basis for continuous moni-
ground and degree of social support, and de- toring of consumer responses to healthcare
mographic variables, such as age. gender, and deliver)'. However, the quality of patient satis-
education, have been found to influence pa- faction with nursing care measures varies con-
tient satisfaction ratings in some studies'-^"" siderably and, until recently, few have under-
but not others."' PasctJe'^ found that older gone rigorous psychometric evaluation. The
patients tended to report higher satisfaction majority of patient satisfaction instruments in
with nursing care, whereas a study by Bader'** the literature are not based on explicit theo-
did not support this finding. Women have retical models. A critical feature of a usable
been found to be more satisfied with care than patient satisfaction with nursing care measure
men in some studies,'^'^ while other studies is its ability to yield actionable results, that is,
have found no gender differences.^"'•^^ specific data that can be used by managers
Patient expectations also have been found to improve patient care quality. Many exist-
to influence satislaction ratings. Swan^ re- ing tools have been criticized for lacking this
ported that patients with lower expectations quality.
and less knowledge of available services were Research has also shown that a valid
more satisfied with their nursing care. In ad- me;isure of patient satisfaction with nurs-
dition, patient health status has been shown ing care quality must include both patient
222 JOURNAL OF NURSING CARE QUALITY/JULY-SEPTEMBER 2005

and provider perspectives, since these often available in the nursing and health services
differ.^^ In a qualitative study of recently hos- literature.** Several criteria were used to as-
pitalized patients, Larrabee and Bolden^'* iden- sess the relative value of these tools. First, the
tified 5 themes of good nursing care from comprehensiveness of the instrument was as-
the patient s perspective: providing for my sessed with regard to factors identified in the
needs, treating me pleasantly and with re- literature as determinants of patient satisfac-
spect, caring about me. being competent tion with nursing care quality. Also, on the ba-
(accurate knowledge and skills), and provid- sis of the recommendations in the literature
ing prompt care. In another study, Larrabee that measures of patient satisfaction should
and colleagues^^ found that aciiievement of reflect both nurse and patient perspectives,
patient-identified go;ils was a strong predic- these tools were further evaluated to deter-
tor of patient satisfaction with nursing care mine the extent to which quality indicators
quality (as measured by a modified version of that the patients identified in the literature
the Nursing and Daily Care Subscale from the were included. The instruments were also
Patient Judgment of Hospital Quality fPJHQ] evaluated in terms of the availability of psy-
questionnaire).^*' Several studies have demon- chometric data, readability of items. length
strated differences between nurses' and pa- of the tool, ease of scoring, and sensitivity
tients' perceptions of important characteris- to actual nursing activities/responsibilities. Fi-
tics of care provided by nurses. Nurses in nally, the instruments were assessed to deter-
several studies consistently overestimated the mine the extent to which the results of the
importance of emotional care for patients in responses to items on the questionnaire were
comparison to patients' perceptions.^^"^" Pa- "actionable," that is, their usefulness to nurs-
tients gave higher ratings to the importance ing administrators for improving the patient
of technical care, such as providing explana- care processes.
tions regarding their condition and care, fol- Few of these instruments addressed all of
lowing through, and monitoring. Congruence the factors found in the literature. On the ba-
between nurse and patient expectations for sis of a review of the literauire and initial
good nursing care are important. Kovner*' feedback from 9 professional practice lead-
found that the less patients and nurses dis- ers on the patient satisfaction questionnaires
agreed on desirability of outcomes, the more currently used in their organizations, we de-
satisfied patients were with their care. Tlius, cided that the PJHQ questionnaire^*^ was po-
it is important for measures of patient satis- tentially useful as a measure of patient sat-
faction with nursing care to include patient- isfaction with nursing care. Modifications to
identified lactors of good nursing care.''^ the tool were made to capture a broader
scope of nursing care activities across the en-
PRELIMINARY WORK tire patient care process from admission to
discharge. The new instrument, tbe Patient
The initial work for this study was a compo- Satisfaction with Nursing Care Quality Ques-
nent of a larger project that evaluated the va- tionnaire (PSNC:QQ), was developed with the
lidity of a variety of nurse-sensitive outcomes intention that this added specificity would
to determine the feasibility of collecting these produce results that would be more action-
data on a regular basis as part of the pa- able and, therefore, able to inform quality
tient discharge abstracting process.'*^ The lit- improvement initiatives. This study was sub-
erature for several nurse-sensitive outcomes, sequently designed to determine the psycho-
including patient satisfaction, was reviewed metric properties of the newly developed
for evidence of their relationship to nursing PSNCQQ, as part of a province-wide initiative
activities. designed to assess the performance of hospi-
The review of the patient satisfaction litera- tals applying a balanced scorecard approach
ture included an evaluation of 29 instruments using the domains of financial performance,
Psychometric Analysis of the Patient Satisfaction Questionnaitv 223

system integration and change, clinical out- daily care, ancillar>' staff and hospital envi-
comes, and patienl satisfaction.'''''" We re- ronment, medical care, information, admis-
port ihc results of this study in the remaining sions, dischai^e and billing, overall quality of
sections of this article. care and services, recommendations and in-
tentions, and overall health outcomes. Previ-
ous Cronbach a reliability coefficients have
METHODS been high, ranging from .90 to .94.^^-ift--^''^«
On the basis of the findings of our critical
A descriptive survey design was used to test review of the literature, items from other sub-
the PSN<:QQ within a random sample of \A scales of the original PJHQ were adapted to re-
hospitals in Ontario. Canada, composed of S flect satisfaction with components of nursing
teaching, 5 large community, and 4 small fa- care. Contrary to Larrabee and colleagues'^^
cilities. All medical and surgical patienls dis- modification of the Nursing and Daily Care
charged in April. May. and June of 2(M)2 from PJHQ subscale, we adapted selected items
these hospitals constituted the study sample. from all 9 PJHQ subscales. The PSNCQQ has
Response rates for similar province-wide re- 19 items, plus .S additional questions designed
search initiatives ranged between M% and to tap satisfaction with the overall quality of
65% for each hospital. On the basis of this care during the hospital stay, overall qual-
past experience, the expected response was ity of nursing care, and intention to recom-
approximately KM) patients per hospital in mend the hospital to family and friends. Each
teaching and large community hospitals. In item of the PSNCQQ consists of a phrase
small hospitals, where fewer patients are dis- to designate the content of tbe question or
charged each month, more than 100 surveys "sign-post," followed by a more detailed ques-
were sent to patients from each small facil- tion or "descriptor." For example, in the first
ity. Respt>nses from patients who visited small item of the instrument, "information you were
hospitals were expected to range trom 60 to given" is used a.s a signpost for the descrip-
90. tor that follows, "How clear and complete the
Patients were eligible if they were older nurses' explanations were about tests, treat-
than 18 years and had not participated in the ments, and what lo expect.'"ITie item format
annual provincial patient satisfaction survey of the PSNCQQ can be seen in Table t.
less than 2 months prior to the study start A S-point Ukert scale ranging from poor
date. To prevent overlap with other pmvin- to excellent is used for each item of the
cial research initiatives underway at the time, PSNCQQ. There are 2 methods to score the
psychiatry and obstetrics patients were also E^NCQQ. For general results, the scores for
excluded. all items can be summed and averaged to
yield a single value for each patient. For de-
Instrumentation tailed feedback and more "actionable" results,
The PSNCQQ was derived from the PJHQ item means and standard deviations can be
questionnaire.^' A multidisciplinary research calculated. Another option is to compute the
team at the Hospital Corporation of Amer- percentage of "strongly agree" responses for
ica developed the original PJIIQ instrument. each item. These results can be used to track
Items were derived from an extensive litera- changes over time or to evaluate the effects
ture review, focus groups, and a content anal- of quality improvement initiatives, llie results
ysis of patients' verbatim ans'wers to questions Cim also be used for comparisons between
about hospital quality'. The goal was to de- units and hospitals.
velop a questionnaire that would represent Prior to testing the new questionnaire, 6
the patient's perspective with content that face-to-face focus groups with nurse stake-
represented salient features to patients. The holders across the province were conducted
original tool contained 9 scales: nursing and to obtain feedback on the new instrument.
224 JOURNAL OF NURSING CARE QUAUTY/JULY-SEPTEMBER 2005

— X
— O -H — ^ —
o o — O •-"

O ON -r,
dv X h^ vq -3 X X
1^, rr. rr.
f-i fT, rn
^ J ^ ^

QC Ov Ov
P X
ON
fO lA O
O Ov ^
add d o
IN ON Ov (N
00
r
q
rr.
/^^. ^ ^
X
ON
X O
G\ O
P ^H - - -H O O — o —
O o —^ — -- p_
§g X vq IA X X
rrt rr. rr.

(N O 00 P rr, P
— O O '— o O

Ov •<!ri X IT
OV O\ r-^ — ON
o O
u rr^ rTi fT, •^ rn

in

•-I

a
o

-3
-3

u
z
Psychometric Analysis of the Patient Satisfaction Questionnaire 225

'The panel participunts felt that the new in- 525) than males (n = 445"): the average length
strument acciinitely reflected what nurses do, of stay was 6.H days. !*atient demographics
and they repeatedly commented on the clarity did not differ significantly by hospital type
of the language used in the instrument. Tliey (Table 2).
also felt that it was advantageous that the in-
strument had already been tested for reliabil- Data analysis
ity and validity. Descriptive statistics and or reliability anal-
yses were conducted. Construct validity was
Data coUectlon established through exploratory factor analy-
Questionnaire packages were mailed to sis and confirmatory factor analysis. After risk-
3,036 patients with a letter of explanation, adjusting for length of stay, gender, age, and
the PSNC;QQ instrument, and a researcher- self-rated health, we examined the extent to
addressed stamped return envelope. Strate- which the PSNCQQ predicted overall satis-
gies suggested by Dillman''' to improve the re- faction with care received during ihe hospi-
turn rate were used in an attempt to maximize tal experience, overall satisfaction with nurs-
the response rate. Three mailings were con- ing care, and intent to recommend the hos-
ducted in September, October, and November pital to famil\ and friends. Subgroup anal)ses
2(K)2. A reminder letter was sent 3 weeks after to examine differences among hospital groups
the first mailing, and a second questionnaire were conducted using analysis of variance
(A'NOVA) procedures.
was mailed 3 weeks later. Completed qties-
tionnaires were returned to the professional
survey org;mization that handled the data col- RESULTS
lection process. Data was entered into an SPSS
data file and cleaned prior to data analysis. Descriptive statistics
Tlie final sample consisted of 1041 patients Tlie means and standard deviations for
(3-4% response rate), with 492 patients from the total P S N ( ; Q Q scale, overall satisfaction
teaching hospitals. 38S patients from large with cart" measures, and intentions to recom-
ct>mmunit>' hospitals, and 164 patients fnim mend the hospital to family or friend sub-
small hospitals. The overall average patient scale are presented in Table 3. Consistent with
age was 63, with slightly more females (« = previous studies found in the literature, the

Table 2. Patient demographics

SD)
Teaching Community Small AU
Age ()'ear5) 61.94(16.92) 63.80(15.70) 63.66(16.18) 62.94(16.35)*
Length of stay (days) 7.34(13.78) 6.71 (9.48) 5.45 (7.08) 6.81 (U.45)'
n (%) «(%) n (%) fi(%)

Gender
Female 269 (54.7) 167 (53.2) 89 (54.3) 525(54.1)^
Male 223 (45.3) 147(46.8) 75 (45.7) 445 (45.9)^
Marital Status
Married 170(57.6) 63 (60.6) 451 (65.3)^

*Nonsipnificani F test. P > .05.

' Noiujgnificant xi i<^i. P > 05
226 JOURNAL OF NURSING CARE QUAIJTY^ULY-SEPTEMBER 2005

Table 3. Means and standard deviations of patient satisfaction scores by hospitai type

3f(SD)
Overall perceptions Teaching Community Small AU

Total Patient Satisfaction with Nursing Care 3.19(0.92) 3.76 (0.88) 3.98 (0.82) 3.81 (0.89)
Quality Questionnaire
Overall quality of care and services you 4.06(1.04) 3.97(1.01) 4.16(1.01) 4.04 (I 02)
received during your hospital stay
Overall quality of nursing care you received 4.04(1.06) 4.03 (0.99) 4.23 (0,96) 4.06 (I 02)
during your hospital stay
On the basis of nursing care I received, I 4.38(1.04) 4.17(1.12) 4.32(1.07) 4.30(1 08)
would recommend this hospital to my
family and friends

responses to all items in the PSNCQQ were suggests that the PSNCQQ is tapping relatively
hij^ily skewed (range was 0.447 to 1.29). consistent levels of patient satisfaction regard-
The majority of patients rated their satisfac- less of hospital t>'pe.
tion as either very good or excellent. Al-
though PSNCQQ mean scores were similar Reliability analyses
across hospital types, patients in lat^e com- The Cronbach a reliability estimates for the
mutiity hospitals were significantly less satis- PSNCQQ were excellent (.97). Item total cor-
fied than patients in smaller hospitals on the relations w^ere high, ranging from 0.61 to 0.89.
total PSNCQQ scale and overall satisfaction Reliability estimates were similar across hos-
with hospital care. Teaching and large com- pital categories (Table 4). This suggests that
munity hospitals did not differ on any satisfac- patients in different types of hospital systems
tion measures. were interpreting the items on the PSNCQQ
To ascertain the source of hospital-type in a consistent manner.
differences in patient satisfaction, w^e used
ANOVA to examine group differences on Validity
PSNCQQ item means (see Table 1). Patients c:onstruct validity was established through
in small hospitals were significantly more sat- exploratory factor analysis and confirmatory
isfied than those in large community hospitals factor analyses. The results of the EFA revealed
on 5 of the 19 PSNCQQ items: willingness of
nurses to answer their questions, the extent to Table 4. PSNCQQ Cronbach a reliability esti-
whieh nurses kept them informed about their mates by hospital type*
condition, nurses' willingness to be flexible
in meeting their needs, and provisions for a
restful environment (P < .05). They also were Item total
correlation
more satisfied than patients in teaching hospi-
Cronbach a range
tals on the following items: how well nurses
kept them informed, how flexible the nurses Teaching .97 .61-.89
were in providing care, and the promptness (A)mmunity .97 ,71-.86
of nurses' responses to their calls for help. Small .97 .64-.84
Although all patient satisfaction item means All .97 .73-.87
were highest in small hospitals, most were not
significantly higher (P > .05), and the actual •PSNCQQ indicates Patient Satisfaction with Nursing
differences among groups were not large. This Care Quality Questionnaire.
Psychometric Analysis of tbe Patient Satis/action Questionnaitv ITl

a 1-faetor solution with factor loadings greater group m = 3.18, SD = 0.59, and M = 1.17,
than 0.70 (range was 0.753 to 0.890). A con- SD = 0.71, for the high and low groups, re-
firmatory factor analysis confirmed the I- spectively). These results demonstrated the
factor model. Various fit indices demonstrated ability of the PSNCQQ to discriminate be-
a good fit of Ihis model to the data (x~ = tween higli and low levels of overall patient
14.56, GFl = 0.944, IFI = 0.958, CFl = 0.958, satisfaction with the care they received during
RMSEA = 0.091) according to standards rec- their hospital stay and added further support
ommended by lientler and Bonett.''" for the construct validity of the itistrument.
The predictive validity of the PSNCQQ was
examined by testing its ability' to predict ex- UMITATIONS
pected outcomes frequently used for valida-
tion purposes in health services research. Af- Owing to funding constraints, we were not
ter adjusting for length of stay, gender, age, able to conduct a second wave of the study
and self-rated health, the PSNCQQ explained to obtain test-rete.st reliability data, which
significant amounts of the variance in the would have strengthened these study results.
oveniil quality of care and services (64%), Although the.sc studies are often plagued with
ovenili quality' of nursing care (73%)- and in- methodological problems, such as poor re-
tent to recommend the hospital to family and turn rates and the failure to obtain follow-up
friends, in the combined sample of hospi- data for all subjects, particularly with patient
tals (55%) (Table 5). Tliese results provide satisfaction surveys, it would have been de-
strong support for the predictive validity of sirable to have a measure of stability for the
the PSNC:QQ. PSNCQQ. This should be addressed in ftiture
To measure the sensitivity of the PSNCQQ, research.
the overall satisfaction with care measure was
dichotomized into 2 groups: excellent/very DISCUSSION AND IMPUCATIONS
good and poor/fair responses. The excel-
lent/very good group had significantly higher The results of the study provide encourag-
scores on the PSNt^QQ than had the poor/foir ing support for the PSNCQQ. The PSNCQQ

Table 5. Risk-adjusted final multiple regression models

Outcome Teaching Community Small All hospitals

Overall quality of care and services
Age. sex, LOS,^ health at admission 0.041 0.010 0.058 0.017
PSNCQQ 0.628* 0.628- 0.574' 0.640*
Final model 0.669' 0.668* 0.611* 0.657
Overall quality of nursing care
Age, sex, LOS, health at admission 0.035 0.004 0.043 0.016
PSNCQQ 0.747' 0.703* 0.668* 0.751"
Final model 0.782* 0.706' 0.71 r 0.746*
Intent to rec<)mmend the hospital
Age, sex, LOS, health at admission 0.025 0.016 0.022 0.018
PSNCQQ 0.585* 0.512* 0.586- 0.552*
Final model 0.608* 0.528* 0608* 0.570'

• Significant at P < .001.

indicates length of stay.
228 JOURNAL OF NURSING CARE QUAUTY/^ULY-SEPTEMBER 2005

showed excellent psychomeiric properties ing). Thus, hospitals could compare their re-
and a strong relationship with overall satisfac- sults to the average scores across other simi-
tion with the quality of care received during lar hospitals in the sample. This feedback was
hospitalization. This is consistent with find- received veryfavorablyby the nurse adminis-
ings in previous research.''^ The PSNC^QQ trators in our sample.
was also strongly related to the likelihood The PSNCQQ can be incorporated into an
of recommending the hospital to friends existing hospital quality monitoring system to
and family. This is particularly important in monitor patient satisfactioti over time. It is a
maritet-driven healthcare systems, such as feasible instrument for use at the unit level
that in the United States. Patient satisfaction owing to its short length and the specificity
with care is a key component of hospital re- of each item to a broad range of nursing care
port cards that are shared with hoth accredi- activities. The instrument could be used on
tation agencies and the public. Most hospitals an ongoing basis to track variabilirj' in patient
use this information as part of their market- satisfaction with nursing care comparatively
ing strategy. In socialized healthcare systems, between units and organizations. Finally, the
the results are often reported in yearly sys- PSNCQQ could also be used as an outcome
temwide report cards and shared with both variable to evaluate the impact of unit or or-
government and the public as part of the pulv ganizational change on patient .satisfaction (ie,
lic accountability process. changes in staff mix, nurse-patient ratios, and
The results of this study yielded actionable, other organizational changes). Tliis informa-
patient-f(K-used results that can be used by tion could be used by nursing management
managers to address areas requiring improve- as one evidence-based indicator of nursing's
ment. Tlie specific nature of the items in the contribution to the patient care process.
new instrument make it possible for man-
agers to identify strengths and weaknesses of CONCLUSION
their nursing care delivery processes. Since
the items are based on fectors identified by The PSNC:QQ appears to have very good
patients as important elements of satisfaction psychometric properties, is of reasonable
with nursing care, actions to address areas of length, is easy to administer, and yields data
improvement are more likely to be appropri- that can be used by management in quality
ate to patients. In our study, each hospital was improvement efforts. Further testing must be
provided with a summary' report of the hos- done to validate the results of this study and
pital means and standard deviations for each add to the database of psychometric prop-
item on the questionnaire, as well as the per- erties for the new instrument. However, the
centage of patients rating each item as excel- results of this study provide encouraging sup-
lent. For comparative purposes, we provided port for the reliability, validity, and clinical util-
the combined data tor all other hospitals of ity of the PSNCQQ as a measure of patient sat-
similar type (ie, small, community, and teach- isfaction that is patient-centered.

REFERENCES

1. <;;)rt-y Rti. l*ijsavac ly. INing patii-nt inturmatton to satisfatlion; amullianalytit appntacli ijtuilKei- Htill.
identify area.s for service improvement. Health Care 19«7;13<4):122-I5().
Manage Rev. 1982:7(2):43-48. 4. HillJ. ftJlicnl satisfaction in a nursc-lcd rheumatolog)-
2. Hays RD, Nelson EC, Rubin HR. Wartr JE. Meterko M. KWrati.J Adv Nurs. l997;2S:347-554.
Further evaluations of the PJHQ scales. Med Care. 5. Under-Pclz .S. Toward a thcorv of patient satisfaction.
I99O;28 (suppl 9);S29-S.19. StKSd.Med. 1982:l(>:5r7-S82.
3. Abfiimowiiz S, CMIC AA, Berry E. Anah'zing patieai 6. Ludwig-Beymer R Ryan C), Johnson NJ, Hennessey
Psychometric Analysis of the Patient Satisfaction Questionnaire 229

KA. Gattuso MC, Epsom R. Using patient percep- 26. Leiter MP Harvie P, Frizzetl C. The correspondence of
tions to improve quality care, J Nurs Care Qual. patient satisfaction and nurse burnout, .Soc Sci Med.
1993;7(2):42-5l. 1998;47(IO):I611-I617.
7. Petersen MB. Measuring patient satisfaction: collect- 27. McCiillis Hall I,, Doran D, Baker R, et al. Nurse staffing
ing useful datay Miz-s Qual Assur. 1988;2(3):2'i-3'>. models as predictors of patient outcomes, Med Care.
H. Risser N. Development of an instrument to measure 2003:41(9):I096-I109,
patient satisfaclion with nurses and nursing care in 28. Swan JE, Carroll MG. Pjtient satisfaction and
primary care settings. Nurs Res. iy75;2i(I):'iS-S2. overview of research—1965 to 1978. In: Hunt HK,
9. Swan J. Deepening the understanding of hospital pa- Day RL eds. Refining Concepts and Measures of
tient satisfoction: fulfillment and equity of effects. Consumer Satisfaction and Complaining Behavior.
J Health Care Mark. I98S;^(3):14. Bloomington, IN: Foundation for the School of Busi-
10. Eriksen LR. P.itient satisfaction with nursing care: ness. Indiana University: 1980:112-118.
concept satisfactioay .'V«r,s Mecis. 199'>;3:5y-76. 29. Weisman CS, Natliason CA. Professional satisfaction
11. Clear>' PD, McNeil BJ. Patient satisfaction as an indi- and client outcomes: a comparative oi^anizational
cator of quality care. Inquiry. 19SH;25;2'>-36. analysis. Med Can: 1985;23(10):l 179-1192,
12. Woodside AG, Frey LL, Daly RT. linking service qual- 30. Rubin HR, Gandeck B, Rogers WH, Kosinski M,
ity customer satisfaction, and behavioral intention. McHorney CA, Ware JE. Patients nitings of outpa-
/ Health Care Mark. 1989:9(4):5-17. tient visits in different practice settings, results from
13. Yellen E. The influence of nurse-sensitive variables the medical outcomes fHuOy. JAMA. 1988;270:835-
on patient satisfaction. AORNJ. 2(X)3;78(5):783. 840.
14. Morales L. Reise S, Hays RD. Evaluating the equiva- 31. Otani K, Kiirz RS, Burroughs TE, Waterman B, Recon-
lence of health care ratitigs by whites and Hispanics. sidering models of patient satisfaction and behavioral
Mei/ Care. 2000:38:517-527. intcntiofLs. Health Care Manage Rev. 2000;28(l):7-
15. Conbere K., McGovem P, Kochevar L, Widtfeldt 20.
A. Measuring satisfaction wilh medical case man- 32. Jaipaul KC. Rosenthal CiE. Do hospitals with lower
agement: a quality improvement tool, AAOHN J. mortality have higher patient satisfaction? A regional
1992:4(K7):333-Ml. 358-360. analysis of patients with medical diagnoses. Am J
16. Rubin HR, Ware JE, Hays RD, The PJHQ question- MedQual. 2OO3;18(2):59 65.
naire. Exploratory factor analysis and empirical scale 33. Lynn MR, Moore K. Relationship between traditional
construction. Med Caw. 1990:28(9 suppl):S22-S29. qtiality indicators and perceptions of care. Semin
17. Pascoe GC. Patient satisfaction in primary health Nurse Manag. 1997:5(4):]87-193.
care: a literature review and analysis, llval Prog 34. Larrahee JH, Bolden LV, I>efining patient-perctrived
Plattn- 1983:6:185-210. quality' of nursing care. J Nurs Care Qual. 20()I;
18. Bader MM. Nursing care behaviors that predict pa- 16(0:34-60,74-75.
tient satisfactitin. J Nurs Qual As.sur. 1988:2(3): 11 - 35. Urrahee JH, Engle VF, Tolley EA, Predictors ol
17. patient-perceived quality. Scand J Caring Sd.
19. Ware J Jr, Davies-Avery A, Stewart A. The measure- 1995;9:153-164,
ment and meaning of patient satisfaction. Health 36. Meterko M, Nelson EC, Rubin HR, et al. Patients' judg-
Med Care Serr Rer. 1978:1:1. ment of hospital quality: a report on a pilot study.
20. Doering ER. Factors influencing inpatient satis&c- Med Catv. 1990:28(suppl):Sl-S56.
tion with care. Qual Rev Bull. 1983;9<:iO):291- 37. Farrell GA. How accurately do nurses perceive pa-
299, tients' needsl' A comparison of general and psychi-
21. Hall JA, Dornan MC. Patient sociodemogntphic char- atric setting.y^rfr N//^. 1991;16:l062-1070.
acteristics as predictors of satisfoctlon with medical 38. von Essen L, Sjoden K). Patient and staff percep-
care: a meta-analysis. Soc Sd Med. 199O;3OC7):811- tions of caring: review and replication, J Adv Nurs.
818. 1991;l6:1363-i374,
22. Sitzia J, Wood N. Patient satisfoction: a review of is- 39. von Essen L, Sjoden PO. Perceived occurrence and
sues and concepts. Soc Sci Med. 1997;45(i2):1829- importance of caring behaviours among patients and
1843 staff in psychiatric, medical and surgical care. J Adv
23. Cleary PD. Rdgman-Lcvitan S, Roberts M. et al. Pa- Nurs. 1995:21:266-276,
tients evaluate their hospital care: a national survey. 40. Young WIJ, Minnick AF, Marcantonio R. How wide is
Health Aff. 1991; 10:25 i-267. the gap in defining quality care? (;omparison of pa-
24. Cleary PD, Keroy L, Kanipanos G, McMullen W Pa- tient and nurse perceptions of important aspects of
tient assessments of hospital care. Qtml Rev Bull. patient care./O/V4. 1996;26:15-20.
1989;15:172-179. 4!. KovnerC, Nurse-patient agreement and outcomes af-
25. Atkins PM. Marshall BS, Javalgi RG. Happy employ- ter surgery West J NuKi Re.'!. 1989;11(1):7-17.
ees lead to loyal patients. J Health Care Mark. 42. Tramner J. Who knows best: the patient or tlie
1996;l6(4):15-23. provider. Hasp Q. 2OOO:3(4):25-29.
230 JOURNAL OF NURSING CARE QUAUTY/^ULY-SEPTEMBER 2005

43. Pringle DM, White P. Nursing matters: the nursing 47. McGiliis Hall L, Irvine Doran I). Laschinger H,
and health outcomes project. Can J Nurs Res. Mallette C, Pedersen C, O'Brien Pallas L. Nursing Re-
port 2001. (Ftvlirnimiiy Study for Hospital Report
44. Laschinner H, Almost J. Patient satisfaction. In: D o 2001). Townto, Ontario, Canada: Fat:ulr>' of Nursing,
nm D, ed. Nursing Sensitive Oulcomes: State of the University of Toronto and Onlario Hospital Associa-
Science. Toronto, Ontario, Canada: Jones & Bartlett tion/Ontario Ministry of Health and Ij)ng-Terni Care;
PiibUshers Inc; 2003:243-282. 2002. LSBN 0-7727 3604-9.
45. Kapian RS, Norton DP The Balanced Scorecard. 48. Nelson EC, Hays RD, Larson C, Batalden PB. Ilie pa-
Translating Strategy into Action. Boston: Harvard tient judgment .system: reliability and validity. Qual
Business School Press; 1996. Rer Bull. 1989;I'5:18S-19!.
46. McCiillis Hail L. Irvine Donin D, Laschinger H, 49. Dillman D. Mail and Telephone Surfeys: The Total
Malleite <:, Pedersen C, OBrien-Pallas L. A bal- Design Method. New York: John Wiley & Sons; 1978.
anced scorecanj approach for Nursing Report 50. Bentler PM, Bonett DG. SiRnificante lests and
Card development. Outcomes Manag Nurs Pract. goodness-ot-fit in ihc analysis t)f covarianee struc-
2(K)3;7(l):17-22. tures. Psychol Bull. 1980;88:588-606.