Patient Satisfaction with the Comprehensive Care Model of Dental Care Delivery

Milieu in Dental Schools and Practice Patient Satisfaction with the Comprehensive Care Model of Dental Care Delivery Ana Karina Mascarenhas, B.D.S., Dr.P.H. Abstract: In the summer of 1997, the College of Dentistry, The Ohio State University, changed its predoctoral clinics from the traditional model to the comprehensive care (CC) model. Although the CC model is considered the better model for delivery of care, from the patient perspective it has not been previously evaluated. The purpose of this study was to compare the two dental care delivery systems the traditional model and the CC model using patient satisfaction. The Dental Satisfaction Questionnaire (DSQ) developed by the Rand Corporation was used to assess patient satisfaction. The questionnaire consists of nineteen items, measuring overall satisfaction and subscales of access, pain management, and quality. The questionnaire was selfadministered to active and recall patients in the summers of 1997 and 1998 to evaluate satisfaction with care in the traditional and CC models respectively. The completed DSQ was returned by 119 respondents in 1997 and 116 respondents in 1998. There were no significant differences in age, gender, and self-rated general and oral health of patients using the two delivery systems. No statistically significant differences were seen in the overall Dental Satisfaction Index and the sub-scales of access, pain management, and quality of care. Statistically significant differences were observed on only two of the nineteen individual items. We conclude that there was no difference in satisfaction levels of our patients between the two dental care delivery models. Dr. Mascarenhas is Associate Professor, Health Policy and Health Services Research, Boston University Goldman School of Dental Medicine. Direct correspondence and requests for reprints to her at Boston University Goldman School of Dental Medicine, Health Policy and Health Services Research, B-306 Robinson Bldg., 715 Albany St., Boston, MA 02118; 617-638- 4456 phone; 617-638-6381 fax; karinam@bu.edu. Key words: patient satisfaction, dental delivery system, dental school, traditional model, comprehensive care model Submitted for publication 3/26/01; accepted 8/28/01 Patient satisfaction with care is a useful measure that evaluates care, including the quality of care and provider-patient relationships. It has been used in medicine for several years, and, as reflected in the recent literature, is increasingly being used in dentistry. 1-5 With the shift in medicine and dentistry to patients being consumers of care and the concept of consumerism, inclusion of patients opinions in assessment of services has gained greater prominence. 6 Patient satisfaction measures the process of care, broadly defined as the professional activities associated with providing care. 7,8 Measuring patient satisfaction allows for evaluation of health systems, particularly comparisons between different models of care delivery. Patient satisfaction is a multidimensional concept. 9,10 Some dimensions of dental care satisfaction that have been identified are technical or aspects of care related to the process of diagnosis and treatment; interpersonal; accessibility/availability; financial access; efficacy/outcomes; continuity of care; facilities; and general or attitudes about overall care. 9,10 Recognizing the need to teach and deliver patient-centered care, The Ohio State University College of Dentistry in 1997 moved its entire predoctoral program from the traditional model of patient care to a comprehensive care (CC) model delivery system. The traditional model of patient care as practiced at The Ohio State University was specialty- or discipline-oriented clinics, with students rotating through these clinics. Patients were referred to these clinics depending on the care they needed. The comprehensive care model of dental delivery is centered around the patient, and is more representative of the model of dentistry practiced in private practice. Patients are assigned to predoctoral students at the initial screening appointment. This student then becomes the patient s primary dental care provider and provides most of the care needed much like the general dentist. If specialty care is needed by the patient, such as periodontal surgery for which the predoctoral student does not have the required expertise, then the patient is referred to the appropriate clinical area with the predoctoral student as- 1266 Journal of Dental Education Volume 65, No. 11

sisting in the patient s care. Advantages of the CC model are that it identifies one primary dental care provider for the patient, there is continuity of care, and the chief complaint of the patient is addressed earlier. Therefore, the CC model is considered the better model for delivery of care both for the clinical training of dentists 11-13 and from the patient care perspective 11 than the traditional model. Although recommended by Gerbert and colleagues 14 in 1996, patient satisfaction with care has not yet been evaluated in the CC model. Patient satisfaction with care is only one of the numerous outcomes that can be measured to assess the effectiveness of the CC model, generally measuring the process of care from the patient perspective. In the broader context, numerous other facets of the comprehensive care model pertinent to the clinical education of dentists as well as the effectiveness and efficiency of care delivery can be measured. Other outcomes that can be measured or have been previously been used as outcomes are number of procedures performed, 11-13 number of patients who receive care, 11,13 overall efficiency of the clinics such as time taken to complete treatment plans, clinic utilization, or total patient charges, 12 and student, faculty, and staff satisfaction with the clinical system. To measure the effectiveness of the CC model in care delivery from the patient perspective, patient satisfaction data was collected in the summer of 1998, a year after the CC model was put in place, and compared to data previously collected in 1997 when the clinic was organized in the traditional model of dental care. The purpose of this paper is to report on patient satisfaction with the two dental delivery systems, traditional model and CC model, and the effect of patient demographics on satisfaction. The null hypothesis was that there is no difference in patient satisfaction between these two dental delivery systems. Methods A review of the literature identified the Dental Satisfaction Questionnaire (DSQ) 9,10 developed by the Rand Corporation as a suitable instrument to be used in the predoctoral clinics. The only change made to the questionnaire was to customize it for our clinics by using the phrases OSU dental students instead of dentists and OSU dental clinics instead of dentist s office. Separate questions on individual patient characteristics such as age and gender, general and oral health, and parking at OSU were added. These questions will not be included in the Dental Satisfaction Index (DSI). The DSQ is a nineteen-item questionnaire, using a 5-point Likert scale ranging from strongly agree to agree, not sure, disagree to strongly disagree. The DSQ measures overall satisfaction with care (DSI) and three dimensions of access, quality, and pain-management. The questionnaire was self-administered to active and recall patients over a three-week period in August 1997 and the same time period in 1998. All active and recall patients during the three-week period were approached to take part in the study in each year of data collection. Two second-year dental students approached patients in the clinic waiting room as they came in for their appointment. The DSQ takes five minutes to complete. Patients were assured that refusal to participate would not impact their treatment. A drop box was used for the patient to return the completed questionnaire. As an incentive to completing the questionnaire respondents were given a toothbrush when they returned it. Approval for this study was received from the Office of Risks and Protection at the Ohio State University prior to its commencement. Statistical analyses were carried out using Epi- Info version 6.0 and the SAS system. The questionnaires were coded and data entered using a customdesigned program in Epi-Info Version 6. Epi-Info was used to compute variables such as the overall DSI score, the subscales of access, pain management, and quality of care, scaled means, and prorated means. The scaled mean is the mean score of the scale divided by the number of items in that scale. 11 The scaled mean puts the overall score, and the subscale scores on a 1 (strongly agree) to 5 (strongly disagree) scoring dimension. Scaled mean scores near 1.0 reflect extreme dissatisfaction, scores near 3.0 reflect neutrality, and score near 5.0 reflect extreme satisfaction. The prorated mean is the raw mean expressed as a percent of the highest possible scale or subscale score. 9 Categorical variables such as self-reported oral health and general health were dichotomized. Epi-Info was also used to obtain descriptive statistics and frequency distributions of all the variables in the dataset. A new dataset was made using Epi-Info so that it could be exported into SAS for further analysis. ANOVA was used to evaluate the difference in mean scores between the traditional and CC models. Chi-square and Fisher s exact tests were used to determine if the observed differences between indi- November 2001 Journal of Dental Education 1267

Table 1. Demographic distribution of the patients in the two clinic systems Traditional Comprehensive Model Care Model (1997) (1998) N=119 N=116 p-value Age 54.9±15.2 53.9±18.7 p=0.66 * Gender Males 41.4% 47.1% p=0.39 Females 58.6% 52.9% Self-Rated General Health Excellent 11.1% 15.7% p=0.83 Very good 35.0% 36.5% Good 38.5% 35.7% Fair 13.7% 12.2% Poor 1.7% 0% Self-Rated Oral Health Excellent 10.3% 5.2% p=0.45 Very good 17.1% 24.1% Good 43.6% 39.7% Fair 22.2% 24.1% Poor 6.8% 6.9% Missing Teeth 85.5% 77.0% p=0.09 (other than 3 rd molars) Use of Prosthesis Complete Dentures 15.7% 11.6% p=0.06 Partial Dentures 30.4% 15.1% Implants 3.9% 2.3% Fixed bridge work 15.7% 19.8% None 34.3% 51.2% * p-value from ANOVA p-value from Chi-square test p-value from Fisher s Exact test vidual satisfaction questions and the dental delivery systems were statistically significant. The Mantel- Haenszel Chi-square technique was used to adjust for, or to evaluate for, confounding factors such as gender, age, and missing teeth. Finally, the hypothesis that there is no difference in satisfaction between the two delivery systems was tested using linear and ordinal logistic regression simultaneously controlling for gender, age, missing teeth, self-reported general and oral health, and use of an oral prosthesis. Results A total of 235 active or recall patients completed the DSQ questionnaire, with 119 respondents in 1997 and 116 respondents in 1998. In each year, only about six patients declined to answer it. As a consequence, more than 95 percent of all active and recall patients completed the DSQ questionnaire in 1997 and 1998. Table 1 reports the demographic distribution and self-rated general and oral health of the patients treated in the traditional model and the CC model. As seen from the p-values, no significant differences were seen in the patient characteristics attending the two clinic systems. Tables 2 gives the mean, standard deviation (sd), scaled mean, and prorated mean for each subscale and overall DSI for the two dental delivery systems. The values for the mean, scaled mean, and prorated mean for the two delivery systems were very similar, resulting in no statistically significant differences in means for overall DSI (p=0.31) or the subscales of access (p=0.13), pain management (p=0.94), and quality (p=0.67). When patients responses to individual items that comprise the DSQ were compared between the two delivery systems, only two of the nineteen items were statistically significant (Table 3). The first item was people are usually kept waiting a long time when they are at the OSU dental clinic (p=0.02). Higher proportions of patients were more likely to agree to this statement or be unsure in the comprehensive care model (40 percent) compared to the traditional model (21.9 percent). The other item was OSU dental students are able to relieve or cure most dental problems that people have (p=0.02). Lower proportions of patients were more likely to agree with this statement in the comprehensive care model (64.3 percent) compared to the traditional model (82.5 percent). These differences were still significant after controlling for age, gender, self-rated general and Table 2: Mean±sd, scaled mean, and prorated mean for the subscales and overall dental satisfaction index with the traditional and comprehensive care models Traditional Model Comprehensive Care Model Number Scaled Prorated Scaled Prorated Scale of items Mean±sd Mean Mean (%) Mean±sd Mean Mean (%) Access 7 24.1±3.8 3.4 68.9 23.3±3.8 3.3 66.6 Pain management 3 10.8±2.5 3.6 72.0 10.7±2.6 3.6 71.3 Quality 7 28.4±2.9 4.1 81.4 28.2±3.4 4.0 80.6 DSI - overall 19 70.7±7.2 3.7 74.4 69.4±8.5 3.7 73.1 1268 Journal of Dental Education Volume 65, No. 11

Table 3. Results of Fisher s exact chi-square test comparing responses to individual items of the DSQ in the traditional model and CC model DSQ Question p-value 1. There are things about the dental care I receive at the OSU Dental Clinic that could be better. 0.34 2. OSU dental students are very careful to check everything when examining their patients. 0.61 3. The fees at the OSU Dental Clinic are too high. 0.61 4. Sometimes I avoid going to the dentist because it is so painful. 0.57 5. People are usually kept waiting a long time when they are at the OSU Dental Clinic. 0.02* 6. OSU dental students always treat their patients with respect. 0.63 7. One of the reasons I come to the OSU Dental Clinic is because there are not enough dentists in my area. 0.59 8. OSU dental students should do more to reduce pain. 0.97 9. OSU Dental Clinic is very conveniently located. 0.10 10. OSU dental students always avoid unnecessary patient expenses. 0.12 11. OSU dental students aren t as thorough as they should be. 0.96 12. I see the same dental student just about every time I go for dental care. 0.10 13. It s hard to get an appointment at the OSU Dental Clinic for dental care right away. 0.78 14. OSU dental students are able to relieve or cure most dental problems that people have. 0.02* 15. Office hours at the OSU Dental Clinic are good for most people. 0.16 16. Dental students usually explain what they are going to do and how much it will cost before they begin treatment. 0.78 17. Dental students should do more to keep people from having problems with their teeth. 0.27 18. The OSU Dental Clinic is very modern and up to date. 0.07 19. I am not concerned about feeling pain when I go for dental care at OSU Dental Clinic. 0.92 * Statistically significant at the p<0.05 level Table 4. Multiple regression model for overall dental satisfaction index with the two delivery systems Standard Independent Variable Estimate Error p-value Age -0.001 0.04 0.97 Gender (males) -0.66 0.79 0.40 Self-rated oral health -0.91 0.61 0.14 Self-rated general health -1.21 0.72 0.10 Missing teeth (except 3 rd molars) -0.65 1.64 0.69 Traditional versus CC model -1.40 1.21 0.25 oral health, and missing teeth using ordinal logistic regression. No significant differences in overall patient satisfaction and the subscales were seen in the two delivery systems in demographic factors, such as age and gender, and other factors such as missing teeth and type of dental prosthesis used. Multivariate linear regression was used to test the hypothesis that patient satisfaction was different between the two delivery systems, controlling for other factors such as patients age and gender, general and oral health, and missing teeth (Table 4). No statistically significant differences were seen in patient satisfaction between the two dental delivery models (p=0.17). Discussion To be able to easily evaluate differences in patient satisfaction between the two dental care delivery systems, two underlying issues are important: the characteristics of the patient populations using the two systems and the instrument used to measure patient satisfaction. The characteristics of the patient populations using the two systems should be as similar as possible, so that any differences in the outcomes are attributed to the differences in the delivery systems, rather than to underlying patient characteristics. We chose to evaluate patient satisfaction in August, as that is the start of the new clinical year. In August 1997 the CC clinics were just getting started, and in August 1998 they had completed a whole year. Using the same time frame eliminated some of the seasonal variations in patients. Use of only active and recall patients guaranteed that they had been in the system for some time. These steps ensured that the patient populations were similar. Further, as reported in Table 1, there were no statistically significant differences in patient demographics and no differences in general and oral health between the patients accessing the two dental delivery systems. This allows us to conclude that the two patient samples were very similar. Therefore, if any differences in patient satisfaction are seen in this study, they are most likely due to differences in the care delivery systems. After reviewing the literature, we chose to use the DSQ developed by the Rand Corporation for the Health Insurance Experiment as a satisfaction measure in our environment. 10 Although not used or reported in the literature very frequently, this was the November 2001 Journal of Dental Education 1269

only dental satisfaction questionnaire that was developed to measure major dental satisfaction constructs, and the validity and reliability of the DSQ has been tested and reported. 9-11 The DSQ measures overall patient satisfaction and three subscales of access, pain management, and quality of care. No significant differences were seen in the overall patient satisfaction or any of the subscales between the two dental delivery systems. These results indicate that those patients attending the College of Dentistry were equally satisfied with the two delivery systems. Although the CC model has been described or depicted as the better model of dental care delivery, from our results it seems that as far as the patient is concerned there is no advantage to the patient in the dimensions of care measured by the DSQ such as access to care, pain management, or quality of care. One reason for this result could be that generally patient satisfaction scores are on the favorable side of the response midpoint as was the case in this study, therefore leaving little room for improvement. Other possible explanations for the results are that a one-year time period is not sufficient to elicit or register change in patient satisfaction or that the process of care measured by the DSQ did not meaningfully differ between the traditional and CC models. The only items of the DSQ that were significantly different were people are usually kept waiting a long time when they are at the OSU dental clinic (p=0.02) and OSU dental students are able to relieve or cure most dental problems that people have (p=0.02). In both cases, patients were less satisfied with the CC model. Again, the CC model is supposed to improve both these dimensions of care. A possible reason for the longer patient waiting times in the CC model is that there was a learning curve that first year for all involved clinic staff, students, and patients. This learning curve resulted in patients being kept waiting longer than before. The perception that students were not able to relieve or cure dental problems of their patients to the same level they had previously been able to may be a function of the CC model. In our CC model the patient is assigned to a single student who acts as the patient s primary dental care provider. The assigned student may be a new clinical student who does not yet have the expertise to treat the patient s problems. Whereas in the traditional model of dental care, patients are assigned to students depending on the complexity of the condition the patient has and the student s expertise or ability to treat. Demographic factors such as age and gender and factors such as self-rated general and oral health, missing teeth, and use of prosthesis did not influence patient satisfaction. When our patient satisfaction results are compared to two other studies by Davies and Ware and Golletz that used the DSQ, our overall satisfaction and subscale scores are higher and better than reported in these studies. 9,10,15 Several reasons can be given for this finding. One is that the mean ages of patients in the present study were much higher than in the Davies and Ware and Golletz studies. 9,10,15 Previous research on satisfaction with health care has shown that older individuals are more likely to report higher satisfaction than younger individuals. 3,6 Another reason is that in the present study the patients were actively accessing care at the College of Dentistry compared to the population based samples in the Davies and Ware and Golletz studies. Other possible reasons for the differences in results include the dynamics of the dental school clinic environment where patients receive low-cost dental services and thus may have lower expectations or are more easily pleased; alternatively patients may perceive that quality of care is better at an educational institution where students work under very close supervision of highly trained faculty. In conclusion, no differences in overall patient satisfaction and in the subscales of access, pain management, and quality were seen between the two dental care delivery models. We will continue to monitor patient satisfaction in the CC model and use results to make improvements. Acknowledgments We would like to thank Matthew Parker and Rudyard Whipps for their invaluable assistance with data collection. REFERENCES 1. Kress G, Shulman JD. Consumer satisfaction with dental care: where have we been, where are we going? J Am Coll Dent 1997;64(1):9-15. 2. Kress GC. Patient satisfaction with dental care. Den Clin North Am 1998;32:791-802. 3. Newsome PRH, Wright GH. A review of patient satisfaction: 1. Concepts of satisfaction. Br Dent J 1999;186:161-5. 4. Croucher R, Robinson P, Zakrewska JM, Cooper H, Greenwood I. Satisfaction with care of patients attending a dedicated dental clinic: comparisons between 1989 and 1994. Int J STD & AIDS 1997;8:150-3. 1270 Journal of Dental Education Volume 65, No. 11

5. Butters JM, Willis DO. A comparison of patient satisfaction among current and former dental school patients. J Dent Educ 2000;64:409-15. 6. Sitzia J, Wood N. Patient satisfaction: a review of issues and concepts. Soc Sci Med 1997;45:1829-43. 7. Donabedian A. Evaluating the quality of medical care. Milbank Memorial Fund Quarterly 1966;44:166-203. 8. Donabedian A. The definition of quality and approaches to its assessment: explorations in quality assessment and monitoring. Volume I. Ann Arbor, MI: Health Administration Press, 1980. 9. Davies AR, Ware JE. Measuring patient satisfaction with dental care. Soc Sci Med 1981;15A:751-60. 10. Davies AR, Ware JE, Jr. Development of a dental satisfaction questionnaire for the health insurance experiment. Santa Monica, CA: The Rand Corporation, 1982. 11. Evangelidis-Sakellson V. Student productivity under requirement and comprehensive care system. J Dent Educ 1999;63:407-12. 12. Holmes DC, Trombly RM, Garcia LT, Kluender RL, Keith CR. Student productivity in a comprehensive care program without numeric requirements. J Dent Educ 2000;64:745-53. 13. Johnson G. A comprehensive care clinic in Swedish dental undergraduate education: 3-year report. Eur J Dent Educ 1999;3:148-52. 14. Gerbert B, Love CV, Caspers NM. The provider-patient relationship in academic health centers: the movement toward patient-centered care. J Dent Educ 1996;60:961-6. 15. Golletz D, Milgrom P, Mancl L. Dental care satisfaction: the reliability and validity of the DSQ in a low-income population. J Public Health Dent 1995;55:210-7. November 2001 Journal of Dental Education 1271