Effects of Participatory Design on Medical Staff's Attitudes Toward The Computerization of Patient Records
Date of Award
Doctor of Science
Center for Computer-Based Learning
Steven R. Terrell
Hubert L. Rosomoff
George K. Fornshell
Many researchers advocate the cause of end-user participation in systems design as beneficial, but these principles are largely limited to opinions and recommendations. Most participatory design studies encountered in specialty literature lack statistical proof regarding the effectiveness of participatory design as a method causing significant behavioral changes in users. This dissertation details a participatory design experiment conducted in the health care environment, involving patient record users having a diverse occupational background. The majority of previous behavioral research studies concerning computers, conducted with subjects employed in the health care field, traditionally concentrate on a particular professional specialty group.
To ascertain current attitudes toward systems used in computer-based patient records applications of medical staff as a whole, the entire medical staff population (N = 105) employed at a not-for-profit Miami health care facility affiliated with the University of Miami School of Medicine was surveyed using the Medical Staff Questionnaire (MSQ). The 30-item Likert type attitude scale was previously developed and validated by the researcher in an independent pilot study (n = 37). Correlational survey results obtained from 104 respondents (99. 05%) indicated that medical staff members had moderately favorable initial attitudes toward computers. Despite the heterogeneous occupational mix of the medical staff population, no significant differences in attitudes toward computers were detected between the 10 professional specialty groups identified at this site of care, suggesting that an approach concurrently involving diverse medical staff occupational categories in experimental behavioral research can be feasible.
Furthermore, of the ten independent variables uncovered from a review of the literature that were used as covariates (gender, age, education, length of employment in health care, use of patient records, typing/keyboard skills, computer use, computer proficiency, computer training, and personal computer ownership), significant differences in mean survey scores were attributed only to self-assessed computer proficiency (p = .001), and PC ownership (p < .05).
It was shown that the majority of these covariates had no significant effect on factor scores calculated for five independent orthogonally rotated factors obtained in exploratory factor analysis for the MSQ instrument: (1) Involvement with computers (use and participation); (2) Impact of computers; (3) Rewards of using computers; (4) Patient Records; and (5) Ethics (staff displacement and legal responsibility) . While the five factors collectively accounted for 49.6% of the variance in item scores, none of the demographic variables collected adequately predicted overall initial attitudes of medical staff toward the computerization of patient records.
To provide acceptable empirical evidence regarding the effects of participatory design on the attitudes of medical staff, a pretest posttest control group type experimental study involving half of the pretested population (n = 52) was conducted. Posttest scores of an experimental group of medical staff (n = 21), randomly selected to participate in health care information systems design activities over a period of 16 weeks, were compared with posttest scores obtained from a randomly selected control group of medical staff (n = 31) who did not receive any treatment. The differences between survey scores of the two post tested groups were found significant at 2-tailed p < .001, indicating that medical staff who actively participated in systems design teams developed more favorable attitudes toward computers than other staff members who were not involved in this process.
As part of the experiment, a functional computer- based patient records software product that met the specific requirements of management and medical staff at this site of care was developed and implemented with end-user participation and support. Rapid modular prototyping, software reuse, and phased implementation techniques were applied, using source code from an existing Informix-SQL database application, the ESS Rehabilitation Manager from Easter Seal Systems (1988), as the starting point. The general orientation and structure of the original application was transformed from a problem-oriented to a patient-based record.
A new relational model, best suited for a site of care offering patient rehabilitation services, was introduced. The experiment facilitated the process of new technology diffusion in the health care organization, by promoting collaborative strategies intended to positively influence users' attitudes and decrease medical staff's resistance toward computers, which may ultimately benefit the quality of patient care.
Mircea Laurian Halmu. 1994. Effects of Participatory Design on Medical Staff's Attitudes Toward The Computerization of Patient Records. Doctoral dissertation. Nova Southeastern University. Retrieved from NSUWorks, Center for Computer-Based Learning. (562)