USF Libraries
USF Digital Collections

Defining service quality in an outpatient clinic with complex constituency

MISSING IMAGE

Material Information

Title:
Defining service quality in an outpatient clinic with complex constituency
Physical Description:
Book
Language:
English
Creator:
Verma, Swati
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla.
Publication Date:

Subjects

Subjects / Keywords:
Service quality
Outpatient
Health care
Patient perspective
Continuous improvement
Dissertations, Academic -- Industrial Engineering -- Masters -- USF   ( lcsh )
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Abstract:
ABSTRACT: The 2001 Institute of Medicine's (I.O.M.) landmark report, Crossing the Quality Chasm: A New Health System for the 21st Century observes that, "though medical science and technology have advanced at a rapid pace,...the health care delivery system has floundered in its ability to provide consistently high-quality care" (I.O.M. 2001). The report recommended six quality aims for a twenty-first century health care system; one of them being patient-centered care. It explains patient-centered care as "providing care that is respectful of and responsive to individual patient preferences, needs, and values and ensuring that patient values guide all clinical decisions" (I.O.M. 2001). This research is aimed at directly addressing this I.O.M. recommendation and seeks to understand quality care in the context of the I.O.M. guideline which clearly states that to achieve quality "the patient is the source of control of interactions" with the provider system.^ The objectives of this project are: (i) to gain a deeper and clearer understanding of the ways patients as customers of an outpatient clinic evaluate health care providers, and (ii) to determine if varying definitions of service quality exist with in a clinic containing a complex constituency. The project site chosen was the set of outpatient clinics at USF Health that makes for a complex site (e.g. eighty different specialties, outpatient surgical units, practicing and academic environment, multi-disciplinary teams at work involving multiple levels of health care professionals and complex inter-personal relationships) to carry out this research. The formal hypothesis can be stated as follows: H1: There exist identifiable differing classes of patients with varying perceptions of Service Quality in an outpatient setting.^ The subsequent research questions that the research aims to address are that, given that differing patient classes can be identified, do they have an impact on the overall patient-perceived quality and how significant is the impact? The project will contribute to a change in the approach at the clinic from a profession-centered to a patient-centered effort. It will raise the awareness among clinicians about how patients view quality care which can then be integrated into the system, institutionalized over time and thus help them improve their ability to provide quality care as preferred by patients. It will also serve to educate and empower the patients by increasing their participation and strengthening their role as partners with clinicians in a health care system. According to a review of the consumer health literature (Hibbard 2003), patients who collaborate with their health care providers and play an active role in their health care have improved health outcomes.^ It also enables future work in metric identification to promote continuous improvement in care provision. Though the research was conducted at a specific outpatient setting, it will have wider applicability as it can be a model worth emulating more broadly. The study also contributes to the academic literature that clearly indicates that there is a recognized need for more research on the delivery of outpatient care (Hammons 2003). Additionally, the study can be applicable and useful in other environments with complex constituencies (e.g. university classrooms, public transportation and travel industry).
Thesis:
Thesis (M.S.I.E.)--University of South Florida, 2007.
Bibliography:
Includes bibliographical references.
System Details:
System requirements: World Wide Web browser and PDF reader.
System Details:
Mode of access: World Wide Web.
Statement of Responsibility:
by Swati Verma.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 49 pages.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001967347
oclc - 263428586
usfldc doi - E14-SFE0002240
usfldc handle - e14.2240
System ID:
SFS0026558:00001


This item is only available as the following downloads:


Full Text

PAGE 1

Defining Service Quality in an Outpatient Clinic with Complex Constituency by Swati Verma A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science in Industrial Engineering Department of Industrial and Ma nagement Systems Engineering College of Engineering University of South Florida Major Professor: Ki ngsley Reeves, Ph.D. Grisselle Centeno, Ph.D. Qiang Huang, Ph.D. Date of Approval: October 30, 2007 Keywords: service quality, outpatient, hea lth care, patient pers pective, continuous improvement Copyright 2007, Swati Verma

PAGE 2

DEDICATION To my Parents and Parents-in-law for their untiring faith and belief. To my Husband for his constant patience and encouragement. To Dr. Kingsley Reeves and Seena Salyani for providing constant support.

PAGE 3

ACKNOWLEDGEMENTS I would like to thank Dr. Kingsley Reeves for his constant guidance, support, belief, and patience. His mentorship was vital to the comp letion of this work. I also would like to thank the committee members and faculty me mbers of the Department of Industrial Engineering at the University of South Florida for their teac hing and support. I also will like to thank Seena Salyani from USF Health for the constant support provided by him in the course of this research. Last but not the least; I would like to thank my husband, Arka Bhattacharya for his enormous support in every way, at every step.

PAGE 4

TABLE OF CONTENTS LIST OF TABLES ii LIST OF FIGURES iii ABSTRACT iv CHAPTER 1 INTRODUCTION 1 1.1 The Health Care Industry 1 1.2 Objectives and Significance 5 1.3 Thesis Organization 6 CHAPTER 2 LITERATURE REVIEW 7 2.1 Technical Quality vs. Service Quality 8 2.2 Extant Models 9 2.3 Patient-Centered Assessment 12 2.4 Approaches Used 14 CHAPTER 3 METHODOLOGY 16 3.1 General Outline of Methodology 16 3.2 Data Collection 18 3.3 Response Rate and Response Bias 20 3.4 Demographics 21 3.5 Survey Reliability and Validity 22 3.6 Data Analysis 24 3.7 Factor Analysis 25 3.8 Factor Analysis Results 27 3.9 Exploring the Factors 32 3.10 Overall Quality 37 CHAPTER 4 DISCUSSI ON AND CONCLUSION 39 REFERENCES 43 APPENDICES 47 Appendix A: Demographic Charts 48 i

PAGE 5

LIST OF TABLES Table 1 Patient Demographics 23 Table 2 Reliability Statistics 24 Table 3 Communalities 28 Table 4 Rotated Component Matrix 30 Table 5 Total Variance Explained 31 Table 6 Variables in Equation 1 35 Table 7 Variables in Equation 2 36 Table 8 Variables in Equation 3 36 Table 9 Variables in Equation 4 38 ii

PAGE 6

LIST OF FIGURES Figure 1 Scree Plot 32 Figure 2 Males vs. Females Distribution 48 Figure 3 Males vs. Females Without OBG/GYN Data 48 Figure 4 Race Distribution 49 iii

PAGE 7

Defining Service Quality in an Outpatient Clinic with Complex Constituency Swati Verma ABSTRACT The 2001 Institute of Medicines (I.O.M.) landmark report, Crossing the Quality Chasm: A New Health System for the 21 st Century observes that, [ though] medical science and technology have advanced at a rapid pace,...the health care delivery system has floundered in its ability to provide consistently high-quality care (I.O.M. 2001). The report recommended six quality aims for a twenty-f irst century health care system; one of them being patient-centered care. It explains patient-centered care as providing care that is respectful of and responsive to individua l patient preferences, needs, and values and ensuring that patient values guide all clinical decisions (I.O.M. 2001). This research is aimed at directly addressing this I.O.M. recommendation and seeks to understand quality care in the context of the I.O.M. guideline which clearly states that to achieve quality the patient is the source of control of in teractions with the provider system. The objectives of this project are: (i) to gain a deeper an d clearer understanding of the ways patients as customers of an outpatient clinic evaluate health care providers, and (ii) to determine if varying definitions of servi ce quality exist with in a clinic containing a complex constituency. The project site chosen was the set of outpatient clinics at USF iv

PAGE 8

Health that makes for a complex site (e.g. eighty different specialties, outpatient surgical units, practicing and academic environment, multi-disciplinary teams at work involving multiple levels of health care professionals and complex inter-personal relationships) to carry out this research. The formal hypothesis can be stated as follows: H1: There exist identifiable differing classe s of patients with varying perceptions of Service Quality in an outpatient setting. The subsequent research questi ons that the research aims to address are that, given that differing patient classes can be identified, do they have an impact on the overall patientperceived quality and how significant is the impact? The project will contribute to a change in the approach at the clinic from a professioncentered to a patient-ce ntered effort. It will raise the awareness among clinicians about how patients view quality care which can then be integrated into the system, institutionalized over time and thus help them improve their ability to provide quality care as preferred by patients. It will also serve to educate and empower the patients by increasing their participation a nd strengthening their role as pa rtners with clinicians in a health care system. According to a review of the consumer health literature (Hibbard 2003), patients who collaborate with their health care providers and play an active role in v

PAGE 9

their health care have improved health outcomes. It also enables future work in metric identification to promote continuous improvement in care provision. Though the research was conducted at a specif ic outpatient setting, it will have wider applicability as it can be a model wort h emulating more broadly. The study also contributes to the academic literature that clearly indicates that there is a recognized need for more research on the delivery of outpa tient care (Hammons 2003). Additionally, the study can be applicable and useful in other environments with complex constituencies (e.g. university classrooms, public transportation and travel industry). vi

PAGE 10

CHAPTER 1 INTRODUCTION 1.1 The Health Care Industry From an economic perspective, health care se rvices are one of the largest and fastest growing industries in the United Stat es. Yet, the last quarter of the 20 th century has been called an era of Brownian motion in h ealth care (I.O.M. 2001). A study of I.O.M. reports over the years (2000, 2001, and 2004) reve als that the American health care sector, which is valued at $1.6 trillion, is suffering from crises deeply related to quality along with safety, cost and access. The concep t of quality patient care is vital to the health care sector and needs increased attention. In fact, improving health care quality is the focal point of health care reform effort s today and has taken center stage away from cost and access in the US public debate abou t health care in the past several years (Chassin 2002). Conventionally, the health care environment has been perc eived as either inpatient or outpatient. Inpatient care requires the patient to stay at the medical center during the course of treatment as opposed to outpatient care, where the patients are not needed to stay overnight. It should be noted that in medical term inology, the terms outpatient and ambulatory care are often used interchangeably. Ambulatory care is an integral part of the health care system in United States and also currently the fastest growing component of 1

PAGE 11

health services delivery in terms of bot h volumes and revenues. As evidence, the National Hospital Ambulatory Medical Care Survey for 2004 reported an estimated 85 million visits to outpatient hospital clinics in the United States in that year, about 29.5 visits per 100 persons. There has been a growing shift from inpatient to outpatient delivery; procedures that once were performed only on an inpatient basis ar e being increasingly performed in a variety of outpatient settings. Advancements in medical technology and the development of noninvasive or minimally invasi ve surgical and non-surgical procedures have contributed to growth in outpatient ambulatory surgical care (Bernstein 2001). This is clearly indicated by the Outpatient Surgery Trends re port that claims the growth of outpatient surgeries to be explosive, from an es timated 400,000 surgeries in 1984 to 8.3 million in 2000. Today, 65% of all surgical procedures do not involve a hospital stay (Lapetina 2002). Also, the managed care plans like Me dicare and a few Health Maintenance Organizations (HMOs) have evolved their reimbursement polic ies over years to limit the inpatient hospital stay durations, thereby encour aging the use of outpatie nt facilities as an economically practical alternat ive over inpatient ones. This growing trend towards outpatient care de mands a consistent, effective, high-quality patient experience in the outpatient envi ronment. Ironically, though there has been a recognized need for more research on the delivery of outpatient care (Hammons 2003), limited information is available on the efforts to promote quality in outpatient settings (Palmer 1988). Several issues related to the quality of patient care persist in outpatient 2

PAGE 12

settings despite the conti nuing shift to outpatient care In contrast, inpati ent settings have received a fair share of attention in rega rds to quality improveme nt. Literature also suggests that there has been a very limited and slow involveme nt of engineering tools and technologies to aid the improvement of the outpatient health systems in delivering quality patient care. This stands in st ark contrast to the manufacturing sector and also some of the service industries like aviation and telecommunications. One of the primary reasons for this could be the fact that health care is very different from the manufacturing sector and in health care, it is important to consider how patients feel about the processes and service they receive at a health care center. The foremost concern regarding quality care is the confusion that prevails in the literature and in practice regarding how quality is defined in a health care setting. The laborintensive nature of health care and latest advances in health care technologies and clinical management of specific conditions has increased the complexity involved in defining and delivering clear and consistent quality in health care (Nicholls 2000). The conflicting expectations of the myriad stake holders only add to the confusion. There is an overwhelming consensus throughout literature that in health care, th ere is a lack of common definition of quality due to divers e professional groupings and inherent characteristics of health care services (Kogan 1991). Another dimension that makes quality an equivocal and ill-defined concept in health care is the problem that lies in the fact that quality is not a single, homogeneous variable bu t rather a complex construct incorporating values, beliefs, and attitudes of individuals involved in a health care interaction (Gunther 2002). 3

PAGE 13

The Quality Chasm report by I.O.M. espoused patient-centered care as one of the most effective views of health care quality. It reiterates the urgent need for more concentrated, rigorous, and critical attention to the role of the consumer/patient in influencing the organization a nd behavior of the health care system (I.O.M. 2001). The recent trend towards the individualization of care, in which the patient is an active participant in decision-making (Waghorn 1999) is gaining currency. This core tenet of the report forms the basis of our research as we attempt to understand the patient perspective of care in an outpatient setting. A glance through the existing body of resear ch that has emerged advocating patients perceptions regarding outpatient experience indicates that the focus has been on addressing issues like average consultation tim es, patient flow, etc., that can be easily measured while the qualitative aspects of service quality have been continually ignored. This is not to say that such efforts are misp laced but to lay emphasis on the possibility of missing out on certain aspects that might carry equal if not more importance in a patients eye and contribute significantly towards th e efficiency and effectiveness of the care delivery by providers, thereby being significant for a more comprehensive evaluation of quality care. The USF Health Outpatient setting makes for a complex and interesting site to carry out this research due to its unique position as an academic setting coupled with a multi4

PAGE 14

specialty clinical environment. The clinics house 80 different specialties with about 400 doctors and support staff employed there. 1.2 Objectives and Significance The objectives of this project are to gain a deeper and cl earer understanding of the ways patients as customers of an outpatient clinic evaluate health care providers and to determine if varying definitions of service quality exist within a clinic containing a complex constituency. The formal hypothe sis can be stated as follows: H1: There exist identifiable differing classe s of patients with varying perceptions of Service Quality in an outpatient setting. The subsequent research questi ons that the research aims to address are that, given that differing patient classes can be identified, do they have an impact on the overall patientperceived quality and how significant is the impact? The research will contribute to raise awar eness among providers regarding how varying patient classes view the quality of care they receive and help them incorporate patients perceptions into the quality-definition and qu ality-measurement process. It will help clinicians customize care to meet the patient requirements while keeping patient preferences and values at the core of care delivery. It also enables future work in metric identification and definition to promote con tinuous improvement and visibility in care provision. 5

PAGE 15

Allowing patients to play an active role in defining quality care a nd collaborating with providers will educate and empower them to participate in service delivery. Since individual preferences are not always concordant with those of their providers, patients need to be involved in decisions about their ca re if their needs and e xpectations are to be met (McNeil 1981). As Coulter notes, perhaps the greatest difference between the envisioned future system and the present re ality is the role of patients themselves (Coulter 2000). In fact, resear ch reveals that increasing pa tients perceived control over their health may affect their hea lth status positiv ely (Rodin 1986). 1.3 Thesis Organization This thesis is organized as follows. Chapter 2 identifies the most important studies related to quality in the health care sector. Chapte r 3 explains the model and methodologies used for the study. It also describes the survey inst rument used for the research and the data collection methods employed. Chapter 4 discusse s the results and presents the model that emerges from the data analysis of the survey responses obtained. It also discusses the limitations faced by researchers and presents possi ble future research applications of this study. 6

PAGE 16

CHAPTER 2 LITERATURE REVIEW The health care delivery system has cha nged tremendously in recent decades. Quality today is a prevailing purpose rather than a desirable accessory (R. Cullen 2000) and the concept of patient-driven quality care is gaining currency. But while there is a considerable body of scholarly work focusing on evaluation of health services from the perspective of providers and clinicians, th e academic literature available on quality care as perceived and defined by patients is far less. Several important aspects of patients perceptions of quality are still not explored and understood by providers and researchers. We still lack a fair idea of what is vital to patients as they assess quality of health care provided to them. In the following subsections, a brief summary of the quality in health care sector as addressed in literature has been presented. The reviewed articles are classified in the following subsections based upon the two kinds of quality that exist in health care (t echnical vs. service), qua lity as viewed by different stake holders involved, models used in literature to asse ss service quality and the approaches used for the same. 7

PAGE 17

2.1 Technical Quality vs. Service Quality Health care quality in literature has been addr essed as either technical quality or service quality. Researchers define technical quality primarily on the basis of the technical accuracy of the medical diagnoses and proced ures or the conformance to professional specifications while service quality refers to the manner in which the health care service is delivered to the patients (Lam 1997). Pa tients have always been in a dependent position as hospitals or other health care pr oviders have specific technical proficiency (know-how) that can be better evaluated by prac titioners, clinicians and medical experts. Most patients are believed to not possess the knowledge or skill necessary to evaluate the quality of diagnoses or the treatment plan. It is now established th at most patients may never determine whether a diagnosis or prescription was optimal or not. A section of articles reviewed questioned the ability of patients to evaluate clinical quality (also called technical quality), with the conclusion that patients find it difficult to distinguish technical quality from service quality (B lumenthal 1996; Laine 1996; Oswald 1998). It must be noted here that terms service quality perceived quality or functional quality are used interchangeably in health care literature Also, terms technical quality and clinical quality mean the same in health care literature. Health care professionals have less regard for service quality while patients base their evaluation of quality on interpersonal and environmental factors (Lam 1997). Patients ar e most capable of evaluating the service quality aspects and frequently use them as surrogates for assessment of aspects they are unable to evaluate as credibly: the accuracy of diagnoses and efficacy of treatment plans which rather tend to be assumed by patient s based on substantiating evidence (Rodie AR 1999). With substantiating evidence author means, for exam ple, if a practice is a 8

PAGE 18

sanctioned provider for a patients employers medical plan and if the provider has the desired credentials on paper, the patient will us e this substantiating evidence to infer that he/she receives high quality me dical treatment (Rodie AR 1999). While it is widely acknowledged that most patients are not qualified enough to judge technical quality, the fact that their asse ssment of service qual ity by several other dimensions that they value the most, can adve rsely affect the total quality experience for them, is vital in defining qua lity care more comprehensiv ely and cannot be ignored. The literature demonstrates that while technica l quality of providers in most cases is considered satisfactory by patients (Friedman 1986), it is service quality or experiences that add to shape up the patients overall view of quality care that needs to be understood better and explored more intens ively. Keeping this in mind, the aim of this research is to determine and focus on the aspects that patients are most capable of evaluating while they receive and consume care in an outpatient setting. 2.2 Extant Models In a review of selected articles aimed at studying health care attributes, the most frequently quoted model was Donabedians cl assic, industrially derived model that segments quality of health care into three categories: structure, process and outcome (Donabedian 1980). Structure largely deals with the physical facilities and environment in which the care is provided. Process refers to the methods (diagnostic and therapeutic) by which the care is provided. Outcome is de fined as the conse quence of the care provided to the patient. The model and the ca tegorization it propagate s has been widely 9

PAGE 19

used and cited but several attempts at modifica tion of this model suggest that it does not always serve as the most useful framework for organizing the wide array of criteria to be used in judging health care quality (S ofaer 2005). The model views quality from a professionals perspective and several modifica tions imply that health care has found the model lacking to address quality as expected and valued from a pa tients perspective. Another oft-cited model used in health care is SERVQUAL model first introduced by Parasuraman et al. in 1985 and further de veloped by them in 1988 to measure service quality from the customers perspective (Parasuraman 1985; Parasuraman 1988). The model has been borrowed from the bu siness world and initi ally proposed ten determinants of service quality that are impor tant to a customer while evaluating services. The ten dimensions of qualit y as initially proposed by the model were based on a series of focus group sessions and are listed as following: tangibles, reliability, responsiveness, competency, courtesy, communication, credibility, security, access, and understanding (Parasuraman 1985). They later reduced the ten dimensions to five for customers to evaluate service quality as tangiblesthe appearance of the physical facilities and materials related to the service; reliability the ability to perform the service accurately and dependably; responsiveness the willingness to help customers and provide prompt service; assurance the competence of the system and its security, credibility and courtesy; and empathy the ease of access, approachability and effort taken to understand customers requirements. The model works with 22 pairs of items that measure the perceived and expected levels of service in a given service industry. It uses a seven-point Likert-type scale for measuring patients' expectations of ex cellent service and their long10

PAGE 20

term experiences of service businesses with the aim of describing service quality at a given point in time (Hiidenhovi 2002). Though wi dely used, the model has been often criticized because though the collective findings by researchers pr ovide support for the validity, reliability and predictive validity of the scale, the factor-l oading patterns in the original five dimensions are inconsis tent across these studies (Lam 1997). The weaknesses of the SERVQUAL model were late r identified and addressed by Ward et al. in their model (Ward 2005). It has been pointed out over years by re searchers that SERVQUAL model may not present a comprehensive view of the dimensi ons of service quality in the health care environment as health care services tend to be more intensive in provider-consumer interactions, which are vastly different fr om the business world that the SERVQUAL model was developed for (Bowers 1994). Othe r related models pr oposed over time to capture patients perception of quality have been various modifications of the SERVQUAL model. Researchers in the newer models have included some of the dimensions that are derived from the SE RVQUAL model along with their own unique approaches to examine the health care service quality. For example, Bowers et al. added caring and patient outcomes to the five qua lity dimensions proposed by the SERVQUAL model after conducting a patient fo cus group interview (Bowers 1994). Another such recent study was undertaken by Wa rd et al., who proposed an integrated view from previous research to examine the quality dimensions comprising the patientperceived quality in the outpatient setti ng. Based upon the previous literature, they 11

PAGE 21

proposed four patient perceived health care quality dimensions: accessgiving patients timely and affordable access to medical care including items such as appointment scheduling, telephone and Web syst em capabilities, information on test results, and cost and insurance issues; outcomepositively impact ing patient health as function of the care given including items such as change in he alth status, and patient s perspective on the referral process; interaction and communicationsgiving patients the experience of constantly courteous and caring treatment from office workers, providers and other involved staff including items like courtesy of front desk staff and provider, general willingness to help, empathy and billing issues; and the final quality dimension tangiblesproviding the patients with the physical facilities, equipment, personnel, and credentials they expect from a health care provider and includes items such as convenience, impression, and layout of facilities, availa bility of needed medical equipment and devices, as well as the credentials of provider and staff (Ward 2005). 2.3 Patient-Centered Assessment Health care delivery involves myriad stake holders and that incl udes providers, payers, physicians, nurses, staff, and the patients them selves (also, patien ts relatives). While earlier approaches towards care delivery were provider-driven, there is a rapidly growing shift towards patient-centered attitudes towards service delivery and patient-focused quality assessment efforts are gaining currency. In health care, se rvices are consumed when they are produced and hence no matter how elusive or diffi cult it is, patient perception of service quality n eeds to be assessed in all health care organizations (Ford 1997). There have been consistent, if limited, efforts to study and examine patients 12

PAGE 22

viewpoints and definitions of outpatient care quality. Most of these efforts however, are based on a general perspective that the needs of all patients are the same. They fail to acknowledge the variations in needs of patients across diverg ent specialties in a given outpatient setting and the influe nce of patient characteristic s on their assessment of care. These approaches presume that the patients visiting an outpatient facility, irrespective of the kind of care they are seeking, have identic al expectations from providers. Thus, it is questionable to what extent this generic appr oach is appropriate for explaining a patients view and assessment of care and understandi ng what drives those perceptions. It is emphasized by some researchers that quali ty care assessments represent a complex mixture of need and expectations and experience of care (Wilkin D. 1992). Over recent years, patient-centeredness in defining quality has been steadily gaining currency. But there has been considerable conf usion in literature in published definitions of patient-centeredness. And researchers agre e that the lack of a universally agreed definition of patient-centeredness has hampered conceptual and empirical developments (Mead 2000) A comprehensive review of literat ure revealed very few studies that assess whether and how patient characteristics relate to perceptions of care quality. A metaanalysis was carried out by Hall and Dornan to examine the relation of patients sociodemographic characteristics (age, ethnicity, sex, socioeconomic status, marital status and family size) to their satisfaction with me dical care (Hall JA 1990). But patients' perceptions of care quality do not automatica lly equate to patient satisfaction (Attree 2000). Confusion prevails in lite rature regarding the relationship of a patients perception of quality care and patient satisfaction. It is ar gued that service quality perceptions should 13

PAGE 23

be considered as long-term consumer attitudes, while patient satisfaction is referred to as short-term, service-encounter-specific judgmen ts (Taylor 1994). While researchers agree that they are not the same concepts, some tend to think they are related and the nature or the direction of the relationship have not been established (Attree 2000), while others believe that they are separate and unique constructs (Westbrook 1991; Oliver 1993). 2.4 Approaches Used Researchers have adopted differe nt approaches to evaluate patients view of quality that range from unstructured qualitative approach es (Appleton 1993; Fosbinder 1995; Kralik D. 1997) to grounded theory methodology (Strauss 1990; Morse 1996). The former approach depends on pre-determined, idealized criteria to be rated by patients using for example a five-point scale (strongly agree, ag ree, neither disagree nor agree, disagree, strongly disagree). The grounded theory appli cations are qualitative approaches which have used the description of patients expe riences of actual care using semi-structured, informal interviews using open-ended questi ons (Attree 2000). Litera ture also mentions clinimetric approach used in inpatient setti ngs that seeks to eval uate quality care by allowing patients to describe the importance and scope of their own reactions and then grouping them into specific categor ies (Feinstein 1983; Sledge 1997). A literature review published in 2005 (Sofaer 2005) regarding qualitative studies that report how patients define quality identified a limited number of small scale studies (eleven to be exact) in a ge neral health care setting. The me thods used were focus groups and patient interviews to determine patients views. Literature also identifies a few 14

PAGE 24

studies using patient experien ce surveys and patient satisfaction surveys. The surveys most often cited in literature are Picker Surveys and CAHPS surveys. These surveys are rigorous and have been developed on the basi s of research using patients themselves. Their validity and reliability have been established by prior research. 15

PAGE 25

CHAPTER 3 METHODOLOGY 3.1 General Outline of Methodology Data on patient perceptions were obtained from a standardized survey of patients across divergent specialties at outpatient clinics associated with USF Health, a complex outpatient setting with 80 differe nt specialties housed under it. We chose to leverage an existing instrument, a survey that contains qu estions relevant to our study. The specialties to source data from were chosen under the guidance of the health experts in order to obtain a sample set that includes patients acro ss seemingly divergent specialties. Patients were chosen randomly from these representative specialties. The quality model chosen for the study is the one proposed by Ward et al., who have proposed an integrated view from previous proposed models to examine the quality dimensions comprising patient perceived qua lity in an outpatient setting (Ward 2005). This models four health care quality dimens ions include the following: accessgiving patients timely and affordable access to medical care; outcomepositively impacting patient health as a function of the care gi ven; interaction and communicationsgiving patients the experience of constantly courteous and caring treatment from office workers, providers, and other involved staff; and tangi blesproviding the patient with the physical 16

PAGE 26

facilities, equipment, personnel, and credenti als they expect from a health care provider (Ward 2005). The goal here is to determine if identifiable differing classes of patients with varying perceptions of service quality ex ist in an outpatient setting (r efer to H1). We decided to use a two-pronged approach to the research problem. One way to approach the problem was to perform an exploratory factor analysis to let clusters of patients (if they exist) with varying needs emerge from the data collec ted through the survey. The idea was that exploratory analysis will help us to identify any patient-class latent in the original dataset, containing pre-determined classes, wh ile the other approach was to focus on the contrasting groups of patients based upon known differences as in age, gender, patient visit status (established vs. new patient visit), etc. One of the variables in the model is the overa ll patient-perceived quality while the other variables are the broad categories of quality dimensions as proposed by the chosen quality model. We use factor analysis to show us if patient classes are valid and if distinct groups can be formed depending on how similarly (or differently) they behave. We also attempted to trace out new, underl ying factors which may be responsible for these groupings. A further analysis is also undertaken to determine whether and how differences in patient classes have an im pact on the overall patient-perceived quality. 17

PAGE 27

3.2 Data Collection Around 10,000 patients were provided with survey s across six major sp ecialties from the outpatient clinics at USF Health from June 2007 to September 2007. The specialties selected were: Cardiology, Pediatrics, Outp atient Surgery, Obstet rics and Gynecology, Family Medicine, and Ophthalmology. The special ties selected with the consultation of medical experts are six of th e busiest and largest specialties at USF Health. In the beginning of the thesis, we provided the info rmation that the outpa tient setting at USF Health houses eighty different specialties. The six major specialties covered here for research purposes include most of the sub-sp ecialties too. For example, under Pediatrics, sub-specialties like General Pediatrics, Infec tious Disease, and Pulmonary Medicine were included. Similarly, under the main speci alty of Surgery, sub-specialties like Cardiovascular Surgery, General Surg ery, Orthopedic Surgery, Plastic and Reconstructive Surgery, Urology, and Vascular Surgery were included. A survey instrument developed by the Leadersh ip Institute Project Team at USF Health was used to capture the patients responses wi th respect to quality dimensions including access to services, facility, interactions and communicati on with staff and provider (physician), and a final question related to patients overall ratings of quality they received that day. The survey was developed in consultation with faculty and upper level management, all medical experts in their own right, at USF Health. The survey was pilottested in two uniquely different sites (Fam ily Medicine and Surgery at two different campuses of USF Health) in December 2006 to es tablish its validity. The reliability of the survey was established and is di scussed later in the thesis. 18

PAGE 28

Surveys were handed out to patients on-site at the time of their vis it to the clinic. The surveys were included in the patients files by the front desk staff and were handed over to them by the physicians they came to see. To make the procedure fool-proof, when the patients saw the front desk staff before their departure, the front desk staff reminded them about filling out the survey. This mode of implementation was employed for four reasons. First, it sends out a st rong message from the provide rs point of view about the concern to improve the service quality when th e physician requests the patients to rate the service received by them. Sec ond, the reminder by the front de sk ensures that forms are filled out by the patients, provided they are willing to, before they leave the premises of the clinic. Third, the patients were expected to fill out the surveys after the fact, at the conclusion of their visit, and not while wait ing for the physician. Finally, it is believed that this may have helped to reduce the respons e bias (if any) with patients. This point is explained further in the latte r sections of the thesis. Every survey carried the unique patient visit number filled out by the front desk staff as the patients arrived. This unique visit numbe r would link the patient responses to the demographic details of the patients stored in electronic patient-record s database. This was done to ensure the privacy of patients at clinics and to save them from entering the demographic details while they may be in a hu rry to leave the clinic after service. The specially marked on-site drop boxes were placed in conspicuous places in clinics in order to make the patients aware of the survey-proce ss and increase their interest in the process to achieve better response rates. The survey was pilot-tested, revised, and finally 19

PAGE 29

conducted from June 2007 onwards. The front desk staff at each specialty was trained to administer the survey and clinic managers were trained to guide the staff in administering the surveys. The reason and significance of th e survey were conveyed to patients before they were invited to complete them. Each survey carried nine questions based upon the quality dimensions proposed by Ward, lik e access, facility, interactions and communication, and overall quality to be ra ted by patients on a ten-point Likert-scale with end points of strongly agree and str ongly disagree. These surveys were scanned electronically to avoid any tampering or huma n error while recording patient responses. Responses were finally integrated with demo graphic details of patients available through electronic patient-records database. 3.3 Response Rate and Response Bias Surveys were administered for approximate ly a month in each of the six chosen specialties. Approximately 10,000 patients received the survey in all of which 1,726 valid patient responses were received. Response ra tes varied across the specialties. While Ophthalmology saw the maximum response ra te amongst all spec ialties at 47.6%, Cardiology was the one with the minimum re sponse rate of 2.7%. The overall response rate at USF Health for our study stood at 17.9 %. The variation in response rates could have been dependent upon the size and nature of the specialty, the involvement of the physicians and the front desk staff, and/or the willingness of the patients to answer the survey. One possible explanation for the high re sponse rates in certain clinics vs. others could be the higher and more dedicated invol vement of the clinic managers or other administrative staff in overseeing the implemen tation of surveys. Another point that came 20

PAGE 30

to light was that physicians will be less inclined to request a survey from the patients that they have a belief received lesser service qua lity on their visits. This may lead to the response bias creeping in the process, as ir ate patients may not receive surveys. Also, certain physicians expressed resentment at handing out surveys themselves as they did not think it was a part of their job. The res ponse bias was tried to be minimized by asking the front desk staff to remind the patients to fill out the survey before they check out, but that again is dependent on the level of th e involvement of the front desk staff. 3.4 Demographics The demographic details provided by the patient -records database were gender, race, age, and established vs. new patients for the clinic s. Of the tota l respondents, 23% were males while 77% were females. The large percentage of female respondents can be attributed to the inclusion of Obstetrics and Gynecology. Race was another demographic data that patient-records could provide data on. The su rvey set contained responses from patients belonging to the following races: Asian or Pa cific Islander, Black, White, White Hispanic and Others/Unknown. Respondents primarily declin ed to disclose th e racial/e thnicity information. Around 69% of all the patients surveyed were categorized as Unknown. About 24% were White, 6% Black, 1% White Hispanic and a miniscule percentage was Asian or Pacific Islander. The patients survey ed were from all ag e-groups ranging from below 18 to 95 yrs of age. To handle the data, we divided patie nts into following agegroups: A1 (<18), A2 (18-25), A3 (26-35), A4 (36-45), A5 (46-55), A6 (56-65), A7 (6675), A8 (7685), A9 (>86). The largest set of responses was from the age-group A3 (18%), followed by A4 (16%). Most of the respondents were primarily females from the 21

PAGE 31

age-groups 26-35. 72.5% of the valid patient res ponses were from established patients vs. 27.5% of them from new ones. Table 1 ca ptures the demographic details of the respondents surveyed. Graphs based on demogr aphics can be viewed in the Appendix section of the thesis under Appendix A. 3.5 Survey Reliability and Validity In this section we address the reliability of the survey used for the research, that is to say, we determine the answer to th e question, Is the survey meas uring things consistently? Mathematically, reliability is defined as the proportion of va riability in the responses to the survey that is the result of differences in the respondents. That is, answers to a reliable survey will differ because respondents have different opinions, not because the survey is confusing or has multiple interpretations. There are a number of ways to determine the reliability of a survey. Some of the commonl y used methods to measure the reliability are: test-retest, split-halves, and internal consistency. We decided to go with the internal consistency approach that considers the inter-item correlation to provide an estimate of reliabili ty. It was employed because this approach avoids the inherent weaknesses associated with the test-retest and split-halves approaches. A common measure of intern al consistency is Cronbach's alpha. The computation of Cronbach's alpha is based on the number of items on the survey (k) and the ratio of the average inter-item co variance to the average item variance. 22

PAGE 32

It can be computed as following: where N is the number of components (items) and is the average of all (Pearson) correlation coefficients between the components. Table 1: Patient Demographics Demographics Respondents Percentage Male 382 23% Gender Female 1284 77% Asian or Pacific Islander 6 0.4% Black 96 5.8% White 397 23.8% White Hispanic 13 0.8% Race Unknown 1154 69.2% <18 216 13% 18-25 197 12% 26-35 287 18% 36-45 251 16% 46-55 205 12% 56-65 217 13% 66-75 156 9% 76-85 102 6% Age >86 22 1% 23

PAGE 33

It generally increases when th e correlations between the ite ms increase and a reliability coefficient of .70 or higher is considered "acceptable" in most research situations. We performed the internal consistency test in SPSS and obtained the results as shown by the Reliability Statistics Table below (Table 2). This establishes a high overall consistency of the survey instrument used for the research. Table 2: Reliability Statistics Construct N of Items Cronbachs Alpha Scheduling 2 0.818 Interactions and Communication 2 0.801 Wait Times 2 0.810 3.6 Data Analysis The goal here is to determine if identifiable classes of patients with varying perceptions of service quality exist in an outpatient setting (hypothesis H1). Exploratory factor analysis is performed to let the clusters of pa tients (if they exist) with varying perceptions of service quality emerge from the responses collected through surveys. This analysis also helped us identify any patient-class late nt in the original da ta-set, containing predetermined classes. Exploratory factor analys is was used to show us if valid patient classes can be formed depending on how simila rly (or differently) they perceive service quality. The next step was to develop various logistic regression models to determine the relationships (if any exist) among these classe s of patients and the demographic variables 24

PAGE 34

available to us. The idea was to look for a ny statistically significant relationship that emerges and then confirm it using confirmatory factor analysis. We also performed a logistic regression in SPSS to predict the imp act of the factors explored in the factor analysis on the patients perception of overall quality that is assessed by question 9 in the survey used. The dependent variable in the model is the overall patient-perceived quality while the independent variables are the br oad categories of quality dimensions as proposed by the chosen quality model. 3.7 Factor Analysis The traditional statisti cal method used by researchers to attempt to identify underlying variables, or factors that explain the pattern of corre lations within a set of observed variables is factor analysis. It is often used in data reduction to identify a small number of factors that explain most of the variance th at is observed in a much larger number of manifest variables. It requires a large sample size as it is based on the correlation matrix of the variables involved, and correlations us ually need a large sample size before they stabilize. There are many different methods th at can be used to conduct a factor analysis (such as principal components analysis, principal axis factor, maximum likelihood, generalized least s quares, un-weighted least squares). There are also many different types of rotations that can be done after the initial extraction of factors, including orthogonal rotations, such as varimax and equimax, which impose the restriction that the factors cannot be correlated (or are or thogonal to each other). 25

PAGE 35

The method used for our analysis is one of the most common forms of factor analysis: principal components analysis. This method is appropriate for cr eating a typology of variables or reducing attribute space. It seeks a linear combination of variables such that the maximum variance is extracted from the variables. It then removes this variance and seeks a second linear combination which explains the maximum proportion of the remaining variance, and so on. This yields factors which are also sometimes called components. Factor loadings, also called co mponent loadings in case of principal components analysis, are the correlation coeffi cients between the variables and factors. The squared factor loading is the percent of variance in that variable explained by the factor. To get the percent of variance in al l the variables accounted for by each factor, the sum of the squared factor load ings is obtained for that fa ctor and divided by the number of variables. Communality is the squared multiple correlation for the variable as dependent using the factors as predictors. The communality measures the percent of variance in a given variable explained by all th e factors jointly and may be interpreted as the reliability of the indicator Low communalities across the set of variables indicate the variables are little related to each other. If the communality exceeds 1.0, there is a spurious solution, which may refl ect too small a sample or the researcher has too many or too few factors. Communality fo r a variable is computed as the sum of squared factor loadings for that variable. For principal co mponents analysis, the in itial communality will be 1.0 for all variables and all of the variance in the variables will be explained by all of the factors, which will be as many as there are variables. The "extracted" communality is the percent of variance in a given variable explained by th e factors which are extracted, which will usually be fewer than all the possibl e factors, resulting in coefficients less than 26

PAGE 36

one. Communality does not change when rotation is carried out. While factor analysis is widely used for data reduction, it suffers th e disadvantage that th e interpretations are intuitive and hence can lead to more than one interpretation of the same data factored the same way. 3.8 Factor Analysis Results Factor analysis was conducted on correlations (as opposed to covari ances) and hence the large sample size (more than 1,700 survey responses) was a perfect fit. SPSS factor analysis (Extraction Method: principal compone nts analysis using listwise deletion of incomplete cases, Rotation Method: Varimax with Kaiser Normalization) was employed for responses obtained from questions 1 to 4 and questions 6 to 8. Question 1 asked patients to rate if health personnel helped am ply in scheduling their clinic visit. Question 2 asked patients to rate if the informati on provided to them before the visit was appropriate. Based on Wards model these qu estions cover the quality dimension access. Question 3 aimed at patients rating the clean liness and orderliness of the facilities and according to Wards model belongs to the qu ality dimension tangibles. Question 4 asked patients to rate the clinic staff for their friendliness and professionalism. Question 6 addresses quality dimension access as it looks at the waiting times patients spent from checking-in to seeing the doctor. Questions 7 and 8 fall under Wards quality dimension called interactions and communicat ion as providers ask patients to rate if their doctor spent enough time discussing th e problem and explaining treat ment options (question 7) and if they were treated with resp ect during their visi ts (question 8). 27

PAGE 37

There has been a conflict among researchers re garding the use of principal components analysis for the ordinal data. While several au thors claim that only c ontinuous data can be used for the principal components analysis, se veral others reject the claim and use it for Likert-scale data. Ward (Ward 2005) et al. have used the approach in their research studies and so have several other authors. From Table 3 we note that all the seven variab les (questions 1 to 4 and questions 6 to 8) are well represented in the common factor space as expressed by their extracted communalities. As noted earlier, low communalit ies across the set of variables indicate that the variables are little related to each other. Our output does not show any particularly low value. Table 3: Communalities Initial Extraction Scheduling1 1.000 .737 Scheduling2 1.000 .731 Facility 1.000 .686 Staff I n C 1.000 .649 Waiting Times 1.000 .997 Provider I n C 1 1.000 .867 Provider I n C 2 1.000 .879 28

PAGE 38

The purpose of the exploratory factor analysis was to see if any la tent factors emerge from the manifest variables. From Table 4 (Rotated Component Matrix), we note that three distinct factors (components) have been extracted and these are the factors we were seeking to discover the pattern s, if any, in the relationship among variables. Questions 1 to 4 (Scheduling1, Scheduling2, Facility, and Staff Interaction and Communication) load on to Component 1. Questions 7 and 8 (Provi der Interaction and Communication 1 and 2) load on to Component 2 and question 6 (Wa iting Time) loads on to Component 3. The loadings on these three factors are good as seen in Table 4. Th is table contains the rotated factor loadings, which are the correlations be tween the variable and the factor. Since the correlations can have possible values ranging fr om -1 to +1, we decided to use a format subcommand in SPSS to not print any of the co rrelations that are 0.3 or less as they are not meaningful when other fact or loadings are good. This make s the output easier to read by removing the clutter of low correlations. The higher the loading of a given quality dimension to a factor, the greater is its cont ribution to the pattern. No quality dimension overlapped between two factors. Though the factor loadings are good, we have to note that the eigenvalues for two of the factors (Component 2 and Compone nt 3) are less than one. This is evident in Table 5 and the Scre e Plot (Figure 1), bo th obtained in SPSS. Table 5 shows one major factor, one moderate factor and one minor factor. This can be possibly explained by the fact that survey contained a limited number of questions as variables to load on to the factors; hence Component 2 ha s two variables and Component 3 has only one variable associated with it. The numbers of questions in the survey were limited to 9, including the question on overall quality to discourage the patients from avoiding to answer a longer, more time-cons uming survey as well as to prevent them 29

PAGE 39

from filling out unrealistic answers due to po ssible fatigue. Hence it was thought to keep the survey one-page long to increase the response rate and the quality of the responses. In future, however, a few more questions can be added to avoid a similar situation. The factor analysis supports our hypothesis that there exist varying cl asses of patients who perceive service quality differently. Table 4: Rotated Component Matrix Component 1 2 3 Scheduling1 .811 Scheduling2 .804 Facility .783 Staff I n C .716 .350 Waiting Times .935 Provider I n C 1 .864 Provider I n C 2 .883 As we look at the three extracted factors, we deduce that from the time the patients decide to use the services of a health care center until they have been seen by physicians, different classes of patients look at the serv ice they received differently and in three phases of their visits. 30

PAGE 40

The first set of patients gives priority to what we call Environment, which includes how easily they could schedule an appointment, the appropriateness of the information provided to them before the visit, cleanliness and orderliness of the facility, and friendliness of the front desk staff. Table 5: Total Variance Explained Component Initial Eigen values Rotation Sums of Squared Loadings Total % of Variance Cumulative % Total % of Variance Cumulative % Total % of Variance Cumulative % 1 4.036 57.659 57.659 4.036 57.659 57.659 2.667 38.096 38.096 2 .861 12.300 69.960 .861 12.300 69.960 1.838 26.255 64.351 3 .650 9.286 79.246 .650 9.286 79.246 1.043 14.895 79.246 4 .504 7.198 86.444 5 .415 5.923 92.367 6 .289 4.124 96.491 7 .246 3.509 100.000 The second cluster of patients gives importance to the physician/health care practitioners attitude towards patients that includes if th e doctor treated a pati ent with respect, and spent enough time discussing his/her problem an d explaining treatment options. The third group of patients gives the highest priority to waiting times they spent from checking-in 31

PAGE 41

at the front desk to seeing the doctor. Thes e three factors collec tively explain 79% of cumulative variance in the da ta as shown in Table 5. Component Number7 6 5 4 3 2 1 Eigenvalue5 4 3 2 1 0 Scree Plot Figure 1: Scree Plot 3.9 Exploring the Factors Based on the three extr acted components, we conducted a few regression tests to see if any statistically significant pattern emer ges between the factors and the patient characteristics (age, gender, r ace, visit status, specialty patie nt visited) available to us. For example, a certain group of patients that belonged to Component 3 gave priority to the waiting times and we attempted to determin e statistically what patient characteristics (if any) impacted this time-sensitive group th e most. For this purpose we used logistic 32

PAGE 42

regression models that may explain the associ ation. The models are discussed in detail in further sections in this chapter. Logistic regression is a regres sion model used for dichotomous dependent variables, that is to say it is appropriate when the re sponses take on only two possible values representing success/failure (0/1). We used bi nary (or binomial) logistic regression, as our dependent variable was di chotomous while the independen t variables were not of any particular type. A linear regression method models the relationship between a dependent variable Y independent variables X i i = 1,....., p and an error term that is a random variable that represents the error in predicting Y from X. The model can be written as where 0 is the intercept (that represents the value of Y when X = 0), the i s are the respective parameters of independent va riables (regression coefficients) and p is the number of parameters to be estimated in the linear regression. When trying to predict the probability that a case will be classified in to one as opposed to the other of the two categories of the dichotomous dependent vari able, we run into a problem. The problem being that the probability can take only take values be tween zero and one, but the predicted values may be less than zero or greater than one. A step towards solving this problem is to replace the probability that Y=1 with the odds that Y = 1 where odds that Y =1, expressed as odds(Y=1), is the ratio of th e probability that Y =1 to the probability that Y Odds can be expressed as follows: O dds = P / (1-P), where P = the probability that Y=1. Though probabilities and odds are equivalent, working with odds have the 1 33

PAGE 43

advantage that odds can take on any positive value; therefore, they do not have any ceiling restrictions. A further transformation of odds eliminates the floor restrictions by producing a variable, the logit or logodds that varies, in principle, from negative infinity to positive infinity. The natural logarithm of the odds i.e., ln {P/ (1-P)} is called the logit of Y, written as logit (Y). If we use logi t (Y) as our dependent variable, we no longer have the earlier problem that the estimat ed probability may exceed the maximum or minimum possible values of probability. The equation for the relationship between the dependent variable and the indepe ndent variable now becomes, logit(Y) = ppX XX ..............2211 It is important to note here that the probabi lity, the odds, and the logit are three different ways to express exactly the same thing and that logistic regression is almost similar to the linear regression with the added advantage, though, that logit tran sformation of odds allows to limit the dependent va riable to be a 0/1 response. We conducted few regression studies in SPSS to determine if any statistically significant pattern emerges between the fa ctors and the patient characte ristics (age, gender, race, visit status, specialty patient visited) available to us. We hypothesize that age, gender and nature of specialty are the major predictors of timesensitive groups. To that effect, we estimated a regression equation in which the variable question 6 that rated waiting times was the de pendent variable and age-groupss, gender, and specialty functioned as independent va riables. Results of the logit model are 34

PAGE 44

presented in Table 6. The output did not reveal any statistically significant predictor of waiting times. Table 6: Variables in Equation 1 B S.E. Wald df Sig. Exp(B) Step 1(a) SPECIALTY 27.367 5 .000 SPECIALTY(1) -.408 .596 .469 1 .494 .665 SPECIALTY(2) -.853 .481 3.147 1 .076 .426 SPECIALTY(3) .360 .328 1.205 1 .272 1.433 SPECIALTY(4) -.146 .329 .197 1 .657 .864 SPECIALTY(5) .024 .360 .005 1 .946 1.025 Constant .659 .318 4.297 1 .038 1.933 Similarly we tried to determine if any of patient characteristics are major predictors of the group that gives the priority to the practitioners at titude towards patients. The dependent variable chosen in this case were the averag e scores of question 7 and question 8 from the survey that rate the quality dimension in teractions and communication of providers. The independent variables were age, gender, specialty, and visit status (established vs. new). The visit status was included to see if the frequency of interaction with the provider has an impact on the way patients perceive the interactions. Results of the logit model are presented in Table 7. Yet again, the output di d not reveal any stat istically significant predictor of the group that gives the prior ity to the practitione rs disposition towards patients and we had to reject the hypothesis. Next regression model that we tried was to determine if time-sensitive groups of patients are sensitive to any day in the week or th e arrival time of the day. The logit model was created in SPSS and the output obt ained is shown in Table 8. We noticed that two of the 35

PAGE 45

time-periods before 2 PM were statistically significant, though the impacts are not very large. Table 7: Variables in Equation 2 B S.E. Wald df Sig. Exp(B) Step 1(a) SPECIALTY 7.782 3 .051 SPECIALTY(1) -1.175 .783 2.252 1 .133 .309 SPECIALTY(2) -.633 .537 1.390 1 .238 .531 SPECIALTY(3) .184 .621 .088 1 .767 1.202 AGE .009 .005 3.637 1 .057 1.009 GENDER(1) .336 .194 2.985 1 .084 1.399 Established vs. New(1) -.073 .212 .118 1 .731 .930 Constant 1.629 .594 7.527 1 .006 5.100 Table 8: Variables in Equation 3 B S.E. Wald df Sig. Exp(B) Step 1(a) Day 7.305 4 .121 Monday .100 .172 .338 1 .561 1.105 Tuesday -.233 .169 1.904 1 .168 .792 Wednesday -.262 .179 2.129 1 .144 .770 Thursday -.186 .177 1.101 1 .294 .830 Arrival Times 17.940 2 .000 7 AM10:59 AM .558 .133 17.557 1 .000 1.746 11 AM1:59 PM .425 .142 8.934 1 .003 1.530 Constant .484 .162 8.942 1 .003 1.623 The research could not detect any statistica lly significant patter ns and in general, relations were extremely small or not shown at all. This leads us to conclude that the patient variables used are not the major pred ictors of a patients view of quality. 36

PAGE 46

3.10 Overall Quality We estimated a regression equation in which the patients view of overall quality was the dependent variable and the que stions based upon scheduling, facility, interactions and communication of staff, and provider were in dependent variables. The questions in the survey are related to the quality dimensions as proposed by Ward model. Binary logistic regression method was employed in SPSS because the dependent variable (overall quality) is a binary/dichotomous response va riable, the outcome being excellent or poor perception of quality (1/0). Results of the logit model are presented in Table 9. The results show that the independe nt variables like physicians giving respect to patients, physicians taking time out to understand th e problems of patients, waiting times, staff friendliness and professionalism and the information patients received before the visit are significant for patients to rate overall quality they received at the clinic. The output also revealed that the strongest predictor of patients perception of overall quality was the way practitioners treated them even when othe r variables were statistically controlled. As patients increase their rating of practitioners (p hysicians) dealing with them with respect (as expressed by Provider2 variab le in Table 9) by one unit, th e odds are that their overall perception of the service quality increases by a factor of 8, when other variables are controlled. The way practitioners treat the patients has the most impact on patients perception of overall quality. Other significant predictors of overall quality were the waiting times for patients to see the doctor fr om the time of check-in and the friendliness of front desk staff. This revelation should be of prime importance to the health care providers as it indicates that patients view overall qua lity primarily based upon how responsive, respectful and comm unicative the practitioners are to them as they receive 37

PAGE 47

support from the staff and behind-the-scen e systems that come into play while maximizing service quality for the patients. In a linear regression model, the coefficient of determination, R 2 summarizes the proportion of variance in the dependent va riable associated with the predictor (independent) variab les, with larger R 2 values indicating that more of the variation is explained by the model, to a maximum of one The regression model used here accounted for 63.6% of the variation in overall quality scores. Table 9: Variables in Equation 4 Step 5(e) Scheduling 2 .908 .254 12.753 1 .000 2.480 Staff 1.144 .256 20.011 1 .000 3.141 Waiting Times 1.326 .263 25.327 1 .000 3.765 Provider1 1.332 .294 20.513 1 .000 3.788 Provider2 2.090 .308 45.971 1 .000 8.087 Constant -2.394 .242 98.149 1 .000 .091 38

PAGE 48

CHAPTER 4 DISCUSSION AND CONCLUSION The fundamental question that inspired this research effort wa s: Are there differing classes of patients that exis t in a given outpatient settin g that view service quality differently? An outpatient setting was chosen over an inpatient setting for the research purpose for two reasons. First, serv ice quality is identified by literature as more vital to an outpatient setting. Second, most of the research efforts in relation to the health care quality have been concentrated in the inpatient environment and literature identifies the strong need to look at quality in an outpatient setting. This study identifies the call in health care literature for further research e fforts in outpatient care delivery considering the growing shift from inpatient to ou tpatient delivery in recent times. The results of the survey conducted at six major outpatient specialties at USF Health and subsequent data analyses reveal that there ex ist three classes of pa tients who view service quality differently. These three classes of patient s divide the process of care delivery into three phases while expecting excellent service quality from the health care centers. One class of patients gives prime importance to the help they receive d in scheduling their appointment for the visit, the friendliness of the staff and the environment at the clinic. The second cluster of patients gives importance to the physician/health care practitioners attitude towards patien ts, if the doctor treate d the patient with resp ect, and spent enough 39

PAGE 49

time discussing his/her problem and explaini ng treatment options. The third group of patients gives the highest priority to waiting times they spent from checking-in at the front desk to seeing the doctor. These th ree factors collectiv ely explain 79% of cumulative variance in the data. Our next effort was to determine if base d upon these factors any significant patterns emerge across the age, gender, specialty, visit status of patients, etc. For example, are there any specific time-sensitive groups amongst the patients surveyed? The research could not detect any statistic ally significant patterns and in general, relations were extremely small or not shown at all. This lead s us to conclude that the patient variables used are not the major predictors of patients views of quality. In terms of overall quality, th is research establishes that the way practitioners treat the patients has the most impact on patients perceptions of overall quality, followed by waiting times for patients to see the doctor from the time of check-in, and friendliness of front desk staff. This revelation should be of prime importance to the health care providers as it indicates that patients view overall qua lity primarily based upon how responsive, respectful and comm unicative the practitioners are to them as they receive support from the staff, and behind-the-scene systems that come into play while maximizing service quality for the patients. Future research is needed to expand the results of this thesis. Patient socio-demographics most often studied and easily collected ar e age and sex. We inte nded to study it beyond 40

PAGE 50

those patient-background variables and include race, visit status (established vs. new patients), economic status, social class, the ki nd of care patients were seeking (acute vs. chronic), etc., but were partially limited by the information patients were willing to provide (for example, though race was one of our socio-demographic variables, about 70% of respondents refused to provide information regarding the race they belong to) and also by the electronic patient-records database that could not provid e us the details we were looking for. Future researchers will do well to expand the surveys ability to capture more socio-demographic details of the patien ts surveyed. Inclus ion of more patientbackground variables in future studies will give better results to determine patients perception of quality. Doing this will enable health care providers to develop a better understanding of the pati ent characteristics and the role they play in a patients perception of care quality. Yet another limitation of this study was the resp onse biases in the patients that may have crept in the process. While every attempt wa s made during the course of administering the survey to minimize the response bias by letting physicians hand out the survey themselves and then front desk staff reminding the patients to fill out surveys before they check out, we still believe that there were scopes for response bias to creep in. For example, physicians will be less inclined to request a survey from the patients that they believe received lesser service quality on their visit. Hence, irate patients may not have received surveys. The response bias was mi nimized by asking the front desk staff to remind the patients to fill out the survey before they check out, but that again is dependent on the level of the involvement of th e front desk staff. Also, it is possible that 41

PAGE 51

chronically ill patients were not informed about the survey and hence were the nonparticipants. This piece of information co uld not be captured by the survey or the electronic patient-records database. This research can be used as a platform for future work on establishing quality metrics in an outpatient care setting. For example, it ca n pave the way for further research in assessing the usefulness of RF ID in an ambulatory healthcare setting to capture real-time data and promote continuous improvement in care provision. In the course of our research, we noticed that the variable waiting times had the highest standard deviation and variance amongst all the variab les rated through the survey and also the highest zero ratings. These can be more effectively and accurately captured by the use of RFID. Through surveys, it is difficult to estab lish which phase of visit the patient spent most time waiting or how much time did the physician spend with the patient. A regular collection of real-time data through RFID can provide meani ngful information that could serve as a useful tool for impr oving quality on a continuous basis. This will also allow service recovery in a remarkab ly shorter period of time. This research can also be extended to othe r environments with complex constituencies like, university classrooms, public transpor tation and travel industry. For example, providers in travel industry can maximize the service qualit y they offer by studying the customer characteristics and the role these char acteristics may play in customers view of quality. 42

PAGE 52

REFERENCES Appleton, C. (1993). "The art of nursing: the ex periences of patients and nurses." Journal of Advanced Nursing 18: 892-899. Attree, M. (2000). "Patients' and relatives' experiences and perspe ctives of `Good' and `Not so Good' quality care." Journal of Advanced Nursing 33(4): 456-466. Bernstein, A. B., Hing, Esther, Burt, Catharin e, and Hall, Margaret (2001). "Trend Data On Medical Encounters: Tr acking A Moving Target." He alth Affairs 20(2): 58-72. Blumenthal, D., and Epstein, Arno ld M. (1996). "The Role of P hysicians in the Future of Quality Management, Part 6." North E ngland Journal of Medicine 335: 1328-31. Bowers, M. R., Swan, J.E., and Koehler, W.F. (1994). "What Attributes Determine Quality and Satisfaction with Health Care Delivery?" Health Care Management Review 19(4): 49-55. Chassin, M. R. (2002). "Achieving And Sustaining Improved Quality: Lessons From New York State And Cardiac Surg ery." Health Affairs 21(4): 40-51. Coulter, A. (2000). Citizens and patients as pa rtners in health serv ices: An overview of various methodologies to incl ude citizens and patients in decision making., Information and Communication in Health, European Health Forum Gastein Congress Report : 155162. Donabedian, A. (1980). "Explorations in Qu ality Assessment and Monitoring : Volume 1. The Definitions of Quality and Approaches to its Assessment. Ann Arbor, MI: Health Administration Press, 1980." Feinstein, A. R. (1983). "An additional basic science for clinical medicine: IV. The development of clinimetrics." Annals of Internal Medici ne 99(6)843-848). Ford, R. C., Bach, S.A., and Fottler, M. D. (1997). "Methods of Measuring Patient Satisfaction in Health Care Organizations. Health Care Management Review 22(2): 7489. Fosbinder, D. (1995). "Patient perceptions of nursing care: an emerging theory of interpersonal competence." Journa l of Advanced Nursing 26: 1085-1093. 43

PAGE 53

Friedman, E. (1986). "What do consumers r eally want?" Healthcare Forum 2: 419-422. Gunther, M., and Alligood, Martha R. (2002). "A discipline-specific determination of high quality nursing care." Journal of Advanced Nursing 38(4): 353. Hall JA, D. M. (1990). "Patient sociodemogr aphic characteristics as predictors of satisfaction with medical care: a meta-analy sis." Social Science and Medicine 30: 811818. Hammons, T., Piland, N.F., and Small, S.D. (2003). "Ambulatory patie nt safety What we know and need to know." The Journal of Ambulatory Care Management 26: 63-82. Hibbard, J. H. (2003). "Engaging Health Care Consumers to Improve the Quality of Care." Medical Care 41(1: Supplement): 61-70. Hiidenhovi, H., Nojonen, K., and Laippala, P. (2002). "Measurement of outpatients' views of service quality in a Finnish univers ity hospital." Journal of Advanced Nursing 38(1): 59-67. I.O.M. (2001). "Institute of Medicine, Crossing the Quality Chasm: A New Health System for the 21st Century." In I.O.M. Comm ittee on Quality of Healthcare in America. Washington, D.C.: National Academy Press. Kogan, M., Henkel, M., and Spink, M. (1991). "Evaluation of total quality management projects in the NHS, first interim report to the Department of Health, Centre for the Evaluation of Public Policy and Practice", Brunel University, London. Kralik D., K. T. W. K. (1997). "Engageme nt and detachment: understanding patients experiences with nursing." Journa l of Advanced Nursing 26: 399-407. Laine, C., and Davidoff, F. (1996). "Patientcentered medicine: a professional evolution." The Journal of the American Medical Association 275: 152-156. Lam, S. S. K. (1997). "SERV QUAL: A tool for m easuring patients opinions of hospital service quality in Hong Kong." Total Quality Management 8(4): 145-152. Lapetina, E. M., and Armstrong, E. M. (2002) "Preventing errors in the outpatient setting: A tale of three states ." Health Affairs 21(4): 26-39. McNeil, B. J., Weichselbaum, R., and Pa uker, S.G. (1981). "Speech and Survival: Tradeoffs between Quality and Quantity of Li fe in Laryngeal Cancer." The New England Journal of Medicine 305: 982-987. Mead, N., Bower, P. (2000). "Patient-centredne ss: a conceptual framework and review of the emperical literature." Social Science and Medicine 51: 1087-1110. 44

PAGE 54

Morse, J. M., and Field, P.A. (1996). Nursi ng Research: The Application of Qualitative Approaches, Chapman & Hall, London. Nicholls, S., Cullen, R., O'Neill, S., and Ha lligan, A. (2000). "Clinical governance: its origins and its foundations." C linical Performance and Quality Health Care 8(3): 172178. Oliver, R. L. (1993). "Cognitive, Affective, and Attribute Bases of the Satisfaction Response." Journal of Cons umer Research 20: 418-430. Oswald, S. L., Turner, Douglas E., Snipes, Robin L., and Butler, Daniel (1998). "Quality determinants and hospital satisfaction." Marketing Health Services 18(1): 19-22. Palmer, R. H. (1988). "The Challenges a nd Prospects for Quality Assessment and Assurance in Ambulatory Care." Inquiry 25(1): 119-131. Parasuraman, A., Zeithaml, V.A., and Berr y, L.L. (1985). "A Conceptual Model of Service Quality and its Implications for Future Research." Journal of Marketing 49(4): 41-51. Parasuraman, A., Zeithaml, V.A., and Berr y, L.L. (1988). "A Multiple-Item Scale for Measuring Consumer Perceptions of Service Quality." Journal of Retailing 64(1): 12-40. R. Cullen, S. N., & A. Halligan (2000). "R eviewing a service, di scovering the unwritten rules." Clinical Performance and Quality Health Care 8(4): 233-239. Rodie AR, P. L., Crabtree BF, McIlvain HE (1999). "Assessing quality. As pressure mounts for clinics to deliver quality, medical practice blueprints and genograms serve as useful tools." Marketing H ealth Services 19(2): 16-24. Rodin, J. (1986). "Aging and Health: Effects of the Sens e of Control." Science 233: 1271-1276. Sledge, W. H., and Feinstein, A.R. (1997). "A clinimetric approach to the components of the patient-physician relationship." The Jour nal of the American Medical Association 278(23): 2043-2048. Sofaer, S., and Firminger, Kirsten (2005). "Pat ient Perceptions of th e Quality of Health Services." Annual Review of Public Health 26: 513-559. Strauss, A., and Corbin, J. (1990). Basics of Qualitative Resear ch: Grounded Theory, Procedures and Techniques Sage, Newbury Park, CA. 45

PAGE 55

Taylor, S. A. (1994). "Distinguishing Servi ce Quality from Patient Satisfaction in Developing Health Care Marketing Stra tegies." Hospital & Health Service Administration 39: 221-36. Waghorn, A., and Mckee, Martin (1999). "Surgi cal Outpatient Clinic s : Are we allowing enough time?" International Journal for Qu ality in Health Care 11(3): 215-219. Ward, K. F., Rolland, Erik, and Patterson, Raymond A. (2005). "Improving Outpatient Health Care Quality: Understanding the Qual ity Dimensions." Health Care Management Review 30(4): 361-371. Westbrook, R. A., and Oliver, Richard L. (1991). "The Dimensionality of Consumption Emotion Patterns and Consumer Satisfaction." Journal of Consumer Research 18: 84-91. Wilkin D., H. L. D. M. (1992). Measures of Need and Outcome for Primary Care, Oxford University Press, New York. 46

PAGE 56

APPENDICES 47

PAGE 57

Appendix A Demographic Charts Males vs. Females Distribution23% 77% Total Males Total Females Figure 2: Males vs. Females Distribution Males vs. Females Without OBG/GYN Data 43% 57% Males Females Figure 3: Males vs. Female s Without OBG/GYN Data 48

PAGE 58

Appendix A (Continued) Race Distribution0% 6% 24% 1% 69% A SIAN OR PACIFIC ISLANDER BLACK WHITE WHITE HISPANIC UNKNOWN Figure 4: Race Distribution 49


xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001967347
003 fts
005 20081024144848.0
006 m||||e|||d||||||||
007 cr mnu|||uuuuu
008 081024s2007 flu sbm 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0002240
035
(OCoLC)263428586
040
FHM
c FHM
049
FHMM
090
T56 (ONLINE)
1 100
Verma, Swati.
0 245
Defining service quality in an outpatient clinic with complex constituency
h [electronic resource] /
by Swati Verma.
260
[Tampa, Fla.] :
b University of South Florida,
2007.
3 520
ABSTRACT: The 2001 Institute of Medicine's (I.O.M.) landmark report, Crossing the Quality Chasm: A New Health System for the 21st Century observes that, "[though] medical science and technology have advanced at a rapid pace,...the health care delivery system has floundered in its ability to provide consistently high-quality care" (I.O.M. 2001). The report recommended six quality aims for a twenty-first century health care system; one of them being patient-centered care. It explains patient-centered care as "providing care that is respectful of and responsive to individual patient preferences, needs, and values and ensuring that patient values guide all clinical decisions" (I.O.M. 2001). This research is aimed at directly addressing this I.O.M. recommendation and seeks to understand quality care in the context of the I.O.M. guideline which clearly states that to achieve quality "the patient is the source of control of interactions" with the provider system.^ The objectives of this project are: (i) to gain a deeper and clearer understanding of the ways patients as customers of an outpatient clinic evaluate health care providers, and (ii) to determine if varying definitions of service quality exist with in a clinic containing a complex constituency. The project site chosen was the set of outpatient clinics at USF Health that makes for a complex site (e.g. eighty different specialties, outpatient surgical units, practicing and academic environment, multi-disciplinary teams at work involving multiple levels of health care professionals and complex inter-personal relationships) to carry out this research. The formal hypothesis can be stated as follows: H1: There exist identifiable differing classes of patients with varying perceptions of Service Quality in an outpatient setting.^ The subsequent research questions that the research aims to address are that, given that differing patient classes can be identified, do they have an impact on the overall patient-perceived quality and how significant is the impact? The project will contribute to a change in the approach at the clinic from a profession-centered to a patient-centered effort. It will raise the awareness among clinicians about how patients view quality care which can then be integrated into the system, institutionalized over time and thus help them improve their ability to provide quality care as preferred by patients. It will also serve to educate and empower the patients by increasing their participation and strengthening their role as partners with clinicians in a health care system. According to a review of the consumer health literature (Hibbard 2003), patients who collaborate with their health care providers and play an active role in their health care have improved health outcomes.^ It also enables future work in metric identification to promote continuous improvement in care provision. Though the research was conducted at a specific outpatient setting, it will have wider applicability as it can be a model worth emulating more broadly. The study also contributes to the academic literature that clearly indicates that there is a recognized need for more research on the delivery of outpatient care (Hammons 2003). Additionally, the study can be applicable and useful in other environments with complex constituencies (e.g. university classrooms, public transportation and travel industry).
502
Thesis (M.S.I.E.)--University of South Florida, 2007.
504
Includes bibliographical references.
516
Text (Electronic dissertation or thesis) in PDF format.
538
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
500
Title from PDF of title page.
Document formatted into pages; contains 49 pages.
590
Adviser: Kingsley Reeves, Ph.D.
653
Service quality.
Outpatient.
Health care.
Patient perspective.
Continuous improvement.
690
Dissertations, Academic
z USF
x Industrial Engineering
Masters.
773
t USF Electronic Theses and Dissertations.
4 856
u http://digital.lib.usf.edu/?e14.2240