Service quality is a focused evaluation that reflects the
customers perception of elements of service. The SERVQUAL model has been
used across various service industries including hospitals to assess and improve
service quality. In this paper Dr Sarika Chaturvedi, Research Associate,
Foundation for Research in Community Health, documents its use in the Indian
todays world of fierce competition, rendering quality service has become
managements top-most competitive priorities and a key determinant of return
on investment as well as cost reduction. In healthcare organisations, the role
that patients play in defining what quality means is now crucial. Owing to information
asymmetry that characterises patient-provider interactions, although the technical
aspects which form the what of a medical service are difficult for
patients to evaluate, the functional aspects about how services
are delivered form important soft components of service delivery.
Evidence suggests that functional quality is usually the primary determinant
of patients quality perceptions and is the single most important variable
influencing consumers value perceptions, which in turn, affect their intentions
to purchase products or services.
Dr Sarika Chaturvedi
Foundation for Research in Community Health
Service quality is a focused evaluation that reflects the
customers perception of elements of service. Parasuraman, Zeithaml, and
Berry (1988) developed a tool to measure service quality the SERVQUAL.
The SERVQUAL has been tested across a number of service industries and its applicability
to the hospital environment has also been assessed in the Western settings.
However, such evidence from Indian hospital sector, and specifically medium
sized hospitals, is sparse. The present paper attempts to fill this gap by reporting
its use to assess the service quality at a Pune based tertiary care hospital.
Materials and Methods
Study Context: This study has been conducted at a
renowned private multi-speciality hospital in Pune, Western Maharashtra, India.
The hospital is functional since over three decades and an established corporate
group manages the business after its take over since recent three years. The
hospital has a bed strength of 110 beds, employee strength of about 400 and
average bed occupancy of over 70 per cent.
Questionnaire, Development and Structure
The developers of SERVQUAL have suggested that it can
be adapted or supplemented to fit the characteristics or specific research needs
of a particular organisation. Hence, we subjected the scale to a preliminary
evaluation. Inputs were received from senior management personnel and an academician.
The decisions to modify the scale were based on relevancy of the questions to
hospital services and ability of the patients to respond to those without undue
frustration or confusion.
The Gap model based on the SERVQUAL depicted in figure one defines service quality
as the difference between perceptions and expectations. It advocates that as
service providers perceptions are important in design and delivery of
services while those of patients are important in the evaluation of services,
the views of both parties are important if a thorough understanding of service
quality is to be gained. Though the SERVQUAL model considers management in the
provider side, keeping in view the Indian hospital sector, which is characterised
by mosty not very large hospitals with very small management teams, this study
has also used staff members instead of only managers. (Henceforth, management
and staff are together referred to as staffs).
The questionnaire included a section on expectations and another on perceptions.
Each section consisted 20 items. These were derived from the Yousseef et al
modified version of the SERVQUAL. The instrument is added with a section three
on demographics (gender, age, education and income) and a final question on
overall service quality of the hospital to be rated on a five point scale. All
questions were close ended.
The scale used is a five point Likert scale with ends anchored strongly disagree
to strongly agree. Though the original SERVQUAL scale uses a seven point Likert
scale, and 22 items this study has used a five point scale with 20 items as
literature shows no association between the number of items, method of administration
and sample size and the reliability of the instrument.
The presentability of the questionnaire was given due attention. Considering
that the scale has 20 statements related to expectations from excellent hospitals
and another 20 about perceptions about the study hospital, common terms were
used for statements. These terms were used instead of repeating the term for
each of those statements as has been done in previous research. In the expectation
scale the term excellent hospitals will have and personnel
at excellent hospitals will was printed as a common term for the 11 statements
and nine statements following these respectively. Similarly, in the perception
scale the common terms were this hospital has
at this hospital.
Another questionnaire was developed for staffs. It included the same statements
as those in the questionnaire for patients, except that the respondents were
asked to mark patients expectations and perceptions, as understood by
them. The common terms, as described above, hence in the staffs questionnaire
were patients expect excellent hospitals to and patients expect
personnel at excellent hospitals to
The questionnaires were made available in English and Marathi languages after
pilot testing. A constant sum scale to determine relative importance of quality
dimensions was put as a separate section as in the originally designed questionnaire.
However, during the pilot testing it was realised that almost all respondents,
in spite of explanation, marked the importance in percentages instead of from
a total of 100 units as was desired. It was then decided to omit this section
to avoid difficulty in response and also to reduce the length of the questionnaire
and rather use regression analysis to reach the objective.
Sample and data collection: A total of 100 patients who had a stay of at least
two days were voluntarily enrolled in the study on the day of/evening prior
to discharge from the hospital. Patients were requested to fill the responses
on the bedside after ensuring that they were comfortable. Each patient took
about 20 minutes to complete. Of the 100 forms filled, five were found to be
incomplete, and were excluded from analysis. All participants were approached
with respect and researcher followed ethical principles in research. Informed
consent was obtained from each participant.
The questionnaire for staff members was administered during their duty hours.
Staff members included in the study were nurses, doctors- generalists as well
as specialists, front office staff, patient assistants and those from accounts,
marketing, human resource and billing departments. Staffs who have worked for
a minimum of three months at the study hospital were invited to participate
in the study. Each staff member took about 12 minutes to complete the form.
The data collection for staff and patients was carried out simultaneously during
the first quarter of 2009.
Results and Discussion
Patients: Male respondents represented about 57 per
cent of the patients surveyed. The study had 52 per cent patients aged below
40 years. The largest group (25 per cent) being in the 21-30 years age group,
the smallest group (five per cent) was aged below 20 years while the elderly
formed about 12 per cent of respondents. Majority of the patients were educated
up to secondary school. Of all the survey questionnaires completed, 39 patients
(41 per cent) did not state their income and were labelled 'Not Stated'. Excluding
these, majority earned below Rs. 20,000 per month. The average length of stay
of the patients as on the day of the study was four days
Staffs: The staff members interviewed included 26
nurses, nine generalist and 14 specialist doctors, and 11others who were staff
from other departments as mentioned above. The staff members surveyed included
65 per cent females and 35 per cent males. The higher number of female participants
is representative of the hospital industry. Of the staffs interviewed most (81
per cent) were less than 40 years of age. Their average work experience in the
hospital industry was 10.6 years while that at the study hospital was 7.8 years.
Validity and Reliability of SERVQUAL Instrument
Considering the objectives of the present study and the recognised instability
of the dimensionality of SERVQUAL, it was considered necessary to address the
construct validity of the scale. It is noteworthy that in the literature about
SERVQUAL, there is no agreement as to which scores (expectation, perception
or quality gap scores) should be factor analysed and indeed, all three types
of scores have been used in previous research. In the present study the researcher
has adopted Vogels et al(1989) view which suggests that the expectation scores
should be factor analysed to determine the items that should be included in
the service quality dimensions because these scores are not influenced
by possible flaws in the service rendered by various firms in the industry.
Thus, in the present study, SERVQUAL scale was factor-analysed by principal
component analysis in the patients expectation scores. The Statistical
package for Social Sciences (SPSS) was used for data analysis. A rotation procedure
was applied to maximise the correlations of item on a factor. Assuming factors
were uncorrelated, Varimax rotation was utilised and four factors with Eigen
values above one were extracted. To measure the adequacy of the sample for extraction
of the four factors the Kaiser-Mayer-Olkin (KMO) measure was computed. The KMO
value (.890) indicates that the examined data set is highly adequate for factor
analysis. Moreover, the data set was found to be multivariate normal and acceptable
for factor analysis according to Bartletts test of sphericity (p = 0.000)
The Bartletts test of Sphericity compared the correlation matrix to the
identity matrix and showed clearly a significant relationship between the variables,
approximately Chi-Square 990.33, df = 190, p < 0.0001.
Total variance explained (63.039) by these four components exceeds the 60 per
cent threshold usually accepted in social sciences to support the solution.
The first factor, which explained 24.37 per cent of the total variance, was
labelled - The human aspect of the service quality. Factor one contains nine
items similar in nature to assurance and empathy and hence could be regarded
as the soft dimension of quality. The second factor includes four
items and explained 13.82 per cent of the total variation. It was labelled
Responsiveness dimension of service quality. Factor three that includes five
items, explained 12.73 per cent of the total variance and was named 'Reliability
dimension of service quality'. The fourth factor comprises three items and explained
12.1 per cent of the variance, it was named 'Tangible dimension of service quality'.
The extracted factors with factor loadings are presented in table one.
The current research results highlighted that the structure proposed by Parasuraman
et al., (1988) for the SERVQUAL scale was not confirmed. This finding is in
line with previous relevant studies. Many of the items loaded heavily into different
factors from the prior dimensions proposed by Parasuraman et.al. (1988). It
was decided to keep these dimensions and analyse the data accordingly. The validity
of the dimensionality of these groups supports the suggestions made by Babakus,
Cronin and others that the dimensions of SERVQUAL may depend on the type of
industry being studied.
An internal consistency analysis was performed to assess the reliability aspect
of the derived four dimensions. The value of the alpha coefficient ranged from
.74 to .89 ( alpha > .70 (Table two) indicating that the four dimensions
are reliable measures of service quality. Reliability analysis was similarly
conducted for the expectation scale and for the perception scale. Both scales
were found to be reliable with Cronbachs alpha value of .92 and .93 respectively.
Patients expectations (PE): In terms of patients
expectation, the mean ranged between 3.73 and 4.60. The lowest 'expectation
score' was for the statement stating Excellent hospitals will have pamphlets
and other communication material visually appealing, while the highest
score was for that stating Excellent hospitals will have the patients
best interest at heart.
This suggests that patients are highly concerned about trust in the hospital.
This could be explained by the mystified nature of medical services or simply
that these are high in credence attributes and hence it is highly difficult
for customers to evaluate them. Another reason for the high expectation could
possibly also be news reports of growing incidences of unethical conduct and
irrational practices in Indian hospitals.
The fact that all the top five expectations are in the human
aspect factor indicate that the management must ensure that the patients realise
that the hospital has patients best interest at heart. It is important
that this is emphasised in communications to patients and also through staff
Amongst the five items that received the lowest expectation scores, three are
from the tangibles dimension while two are from the responsiveness dimension.
Tangible dimension includes items stating about modern looking equipment, visually
appealing physical facilities and visually appealing communication material.
All the tangible dimensions receiving lowest scores indicate that patients do
not go very much by the look of the hospital as is usually assumed.
It is surprising that patients have one of the lowest expectations to staff
never being too busy to respond to patients needs. Possibly patients perceive
a hospital to have a large client base and hence likely to be offering good
quality by noticing staff to be busy.
Patients mean scores for 'perception of actual service' ranged between
3.65 and 4.32. The lowest perception score' was for This hospital
has pamphlets and other communication material visually appealing. The
highest 'perception score' was for the two statements stating The personnel
in this hospital give prompt service to patients and The personnel
in this hospital are always willing to help patients. The findings of
high perceptions in the human factor dimension imply that the personnel are
perceived to be serving well.
The lowest perception is for the hospital has visually appealing communication
material, and about meals being served hot and of good flavour indicating
patients unhappiness about catering services. Indian hospital managers need
to particularly consider this in view of the varied food habits in the country
probably indicating need to give choice of food items to patients.
One item from the human factor that has scored low perceptions is about personnel
telling patients exactly when services will be performed.
Patients perceptions are low about two items from the tangible dimensions.
As these items are also among the low expectation items, the implication is
to include these items in areas of improvement but not in the highest priority
A comparison of patients expectations and perceptions for the four factors
is presented in figure one. Statistical analysis shows that the mean patient
expectations for two of the factors- Factors one and three are significantly
different (p<0.05) from the respective mean patient perceptions.
Staffs Understanding of Patients Expectations and
The mean value for staffs understanding of patients expectations ranged
from 3.87 to 4.70. The lowest score is for statement stating patients
expect excellent hospitals to have visually appealing pamphlets and other communication
material, while highest score was for the statement patients expect
excellent hospitals to always be willing to help patients.
Mean scores for staff understanding of patient perceptions ranged between 3.52
and 4.28, the lowest being for statement five which stated that patients
perceive this hospital has pamphlets and other communication material visually
appealing while highest was for statement which stated that patients
perceive the personnel in this hospital understand the specific needs of their
patients. The mean patient expectation and perception scores as perceived
by staff for each factor are presented in figure two.
Gaps In Service Quality
Gap five: The Customer Gap: This study finds differences
between patients expectations from an excellent hospital and their perceptions
of the service quality delivered at the study hospital. The SERVQUAL model labels
this as gap five- the customer gap. This study finds that there exists gap five
in the hospital analysis reveals that these gaps are significant (p<0.05)
in the human factor and the reliability dimension.
Gap one: The Knowledge Gap: The SERVQUAL model defines
gap one labelled Knowledge gap as the gap between the management/
staff understanding of patients expectations and perceptions and the actual
expectations and perceptions of patients about service quality at the hospital.
It is the first step by which hospitals can proceed to reaching patients expectations.
This entails identifying areas where patients expectations and perceptions of
service quality mismatch with the staff and management understanding of these.
The figure three shows that staff members have largely overestimated patient
expectations. As regards understanding of patient perceptions of service quality,
the reverse is found- staff has underestimated the hospitals performance.
Which Dimensions Matter Most?
In order to examine the effect of the quality gaps - in the four dimensions
- on the patients overall evaluation of the quality of the service provided
by the hospitals (general question in the questionnaire), regression analysis
was performed. The four quality gaps were used as the predictors of overall
quality of the services provided. Considering the independent variables with
statistically significant coefficients, it is evident that patients perceptions
of service quality are attributed to the responsiveness gap as presented in
Table three, which is in fact the predictor of overall service quality. The
above research finding is worth reporting since it indicates that the quality
of the service provided to patients in the study hospital depends heavily on
improving the responsiveness (Factor two).
This study leads to the following conclusions which are particularly important
for further use of the SERVQUAL model in Indian hospital settings.
The SERVQUAL questionnaire can be modified to specific needs as recommended
by Parasuraman et al. However, this raises concern about loss of the power of
standardisation. Although the scale is tested for reliability and validity,
the process of evolution of the scale being subjective, the possibility of negligence
of important items can not be ruled out. The length of the questionnaire is
another important consideration in using the SERVQUAL model. In view of the
middle socio-economic class patients and their cultural contexts, this
study has attempted to improve the presentability and the readability of the
questionnaire which was found useful in keeping participants interest in it.
Involvement of staffs directly interacting with patients instead of management
alone as recommended by the original SERVQUAL model is a unique feature of this
study. This has been found helpful in not only better identifying understanding
of patients expectations and perceptions, but also in creating acceptance
for subsequent service quality improvement strategies.
This study concludes that the dimensional structure of SERVQUAL is unstable
within the hospital industry and this finding is similar to that reported by
Carman (1990) and Babakus and Boller (1992). While the original study by Parsuraman
et al 1988 proposed five (universal) dimensions which were supposed to measure
the service quality in any sector, this study reports four dimensions for the
hospital industry rather than five. This result supports the work of quality
gurus who found that quality is a relative notion with respect to a given client
The regression analysis found the service quality gap in the responsiveness
dimension to be the most strong predictor of overall service quality followed
by reliability. However the model points that there are predictors of service
quality other than the gaps in the four dimensions that this study finds.
Anderson, E.A. Measuring service quality at a university health clinic. Int
J of Health Care Quality Assurance 1995, 8(2), 32-37.
Donnabedian et al. The Quality of Care: How can it be Assessed ,
JAMA 1988, 260(12), 1143-48.
Parasuraman, A., Zeithaml, V.A., & Berry, L.L. SERVQUAL: A multiple-item
scale for measuring consumer perceptions of service quality. Journal of Retailing
1988, 64(1), 12-40.
Youssef, F., Nel, D., & Bovaird, T. Service quality in NHS hospital. Journal
of Management in Medicine 1995, 9(1), 66-74.
Nyeck, S., Morales, M., Ladhari, R., & Pons, F. Ten years of service quality
measurement: reviewing the use of the SERVQUAL instrument. Cuadernos de Difusion
2002, 7(13), 101-7.
Vandamme, R., & Leunis, J. Development of a multiple-item scale for measuring
hospital service quality. International Journal of Service Industry Management
1993, 4(3), 30-49.
Vogels, R., Lemmink, J. and Kasper, H. "Some Methodological Remarks on
the SERVQUAL Model", in Avloniti, G.L. Papavasiliou, NK. And Kouremenos,
A.G. (Eds), Marketing Thought and Practice in the 1990s, Conference Proceedings
of the XVIII Annual Conference of the European Marketing Academy.1989.
Comrey, A. and Lee, H. A First Course in Factor Analysis, Lawrence Erlbaum Associates
1991, Hillsdale. NJ.
Hair, J., Anderson, R., Tathain, R. and Black, W. Multivariate Data Analysis,
4th ed. 1995, Prentice - Hall, Engelwood Cliffs, NJ.
Carman, J.M. Consumer perceptions of service quality: An assessment of the SERVQUAL
dimensions. Journal of Retailing,1990, 66(1), 33-55
Licata, J., Mowen, J. and Chakraboty, G. Diagnosing perceived quality in the
medical service channel, Journal of Health Care Marketing,1995, 15(4), 42-9.
Lim, P.C. and Tag, N.K. A study of patients' expectations and satisfaction in
Singapore hospital, International Journal of Health Care Quality Assurance,2000
. 13 (7), 290-99.
Babakus, E., & Boller, G.W. An empirical assessment of the SERVQUAL scale.
Journal of Business Research,1992, 24, 253-68.,
Cronin, J.J., & Taylor, S.A. Measuring service quality: A re-examination
and extension, Journal of Marketing,1992, 56(3), 56-68.
Nunnaly, J. Psychometric Theory, McGrow-Hill Book Co.1978, New York.