10-2 RADi0LOGY INFORMATION SEARCH SYSTEMS: COMPARISON OF TW O SYSTEMS . - . . . ..L.an.e. Francis Donnelly M.D., C,N.R. Taylor, Childrens H.ospital Medical Center .an.d .... University of Cincinnati, Cincinnati, Ohio . . . . P u r p o s e : Because of increasing restraints on both time and finances, academic radiology departments must maximize efficiency in order to be successful. This study compares the efficiency and accuracy of a commercial radiology support system (RSS) (Ceedata, Cincinnati, Oh) in the management of radiology information with all other methods of searching radiology data available at CHMC. Methods : Two pediatric radiologists were given four identical search tasks, such as "find as many examples of Salter Harris type II fractures as possible". One had access

t o _an_ RSS,._a commercial product attached to our radiology information system ap~ d ...... operating on a direct word search basis, to perform the information searches. Th e other was instructed to use all other methods available. Results: The average time to complete an information search task was 470 hours for the radiologist without access to RSS and .062 hours (3.7 minutes)_for the radiologist with access to RSS. The average number of cases found per task with RSS was 141.5 and without RSS 8.75. Conc lus ion : Radiology information searches were 7,579 times faster and 16.2 times more ,productive when utilizing an RSS. In our opinion, this increased efficiency has facilitated the academic pursuits and administrative abilities of our department.

10-3 RESULTS OF PROCESS IMPROVEMENT AND RE-ENGINEER/NG ON RADIOLOGY REPORT TURNAROUND TIME AT AN ACADEMIC MEDICAL CENTER, Lincoln L. Berland, M.D., J.K. Smith, M.D., Ph.D., University of Alabama at Birmingham, Birmingham, AL ~____qLQ~: In 1994, an initiative to improve radiology report turnaround time was included in ongoing continuous quality improvement efforts. This has led to extensive ongoing operational and itlfrastrucmral changes. Methods: We employed tools and techniques of "Total Quality Management" and "Re-engineering" to effect improvements. Contributions to improvement included increased participation in decision- making by employees, increased staffing of radiology transcription and at-home transcriptionists communicating by modem, periodic nonplmitive feedback to individual radiologist's of their and overall department performance, improved manual and electronic film tracking procedures, reorganization of film filing, revision of film matching procedures, increased immediate interpretation, and estabhshing lknited PACS capabilities. ~ : Between April, 1994 and October, 1995, time from examination completion to electronic report approval on a computerized radiology information system declined from an average of 106.5 hours to 35.2 hours. The greatest improvement occurred in the dine of examination completion to transcription at our outpatient clinic, declining from 126. I hours to 12.8 hours. ~ : Marked improvements in turnaround time can be achieved and sustained. However, improvements can lead to unanticipated problems, and solutions are specific to the environment. Problems must be identified through broad and continuous participation in process improvement, and through attending to both short and long-term objectives.

10-4 ACHIEVEMENT OF SUBSTANTIAL COST REDUCTION THROUGH LARGE GROUP PURCHASING BY THE RADIOLOGY DEPARTMENTS OF A LARGE VERTICALLY ~ R A T E D HEALTH CARE SYSTEM.

Robert Bramson, MD, Brian Chiango, MBA, Steven Seltzer. MD, James Thrall, MD, B. Leonard Holman, MD, Department of Radiology, Harvard Medical School, Partners HealthCare, Boston, MA

Faced with intense pressure to reduce costs, academic medical centers have searched for innovative ways to cut their budgets. We sought to lower the costs by coordination of the purchase of equipment, supplies, and services in the radiology departments of a vertically integrated health system formed by the merger of two of the largest academic medical centers in New England.

Methods: The radiology departments at Brigham & Women's Hospital and Massachusetts General Hospital formed a Cost Reduction Task Force to explore opportunities to jointly decrease expenditures for equipment, supplies, and services. Dam from the operating budgets of both institutions was collected and analyzed to find specific items within the budgets that could yield significant cost savings by joinrefforts.

The pmjecfs first phase yielded over $809,000 in reduced costs from a system-wide annual budget of 7+ million dollars for these items. Ongoing additional projects suggest that longer term contracts, containing steeper discounts, with a decreased number of vendors will result in further decreases in the cost of materials and supplies.

Coordination of purchase by the radiology members of an integrated delivery system can yield substantial.

10-5 DIAGNOSTIC IMAGING IYlILI7_ATION REDUCTION BY PRIMARY CARE PROVIDERS FOLLOWING TF.LEPHONE CAMPAIGN IN A MANAGED CARE ENVIRONMENT, Walid A Hindu M.D. Alan S. Spiro, M.D., Scott Seberg, M.A., Donna Ruggles, ILT.1L, Barbara J. Murphy, R.N., M.A., Alan P. Mintz, M.D., Finch University of Health Sdences/The Chicago Medical School, North Chicago, IL Pu_p_~._~: Reduce unnecessary imaging examinations in management of low back pain. Methods: • Identify physicians with heavy utilization of imaging i~rocedures

• Mail letter outlining guidelines in management of low back pain to these physicians

• Comact these physicians and review material received • Reanalyze data of utilization 6 months after contact

Results: Plans Utilization Reduction Plan A (West) 5496 Plan B (Midwest) 59%

Con.c.tusions: Significant reduction can be achieved in imaging of low back pain with no change in management or outcome with well designed educational efforts. Academic radiologists can play a significant role in improving care and cutting costs in low back pain syndrome and other similar clinical situations with similar strategies..

10-6 UTrI,IZ&TION OF TIME DURING BODY CT IN A TERTIARY CARE TEACHI/qG HOSPITAL: FOCUS ON PATIENT THROUGHPUT, E'nk K. Paul.son. M.D., R.A. Leder, M.D., D.M. DeLong, Ph.D., M.T. Keogan, M.D., A.1L Moore, R.T., K.C. Nelson, M.D., Duke University Medical Center, Durham, NC

~ : To control costs, it is increasingly important to make efficient use of imaging technology. Our purpose was to analyze the time required to complete each step of a body CT scan foousing on factors that influence patient throughput. ~ : Over four weeks, we prospectively monitored the time required for each s t ~ of a body CT scan (image time, check time, and clear time). Covariate dam was collected by patient status [outpatient, inpatient, emergency department fFA)), and iraensive care unit flCU)], work shift, and radiologist training level. Todmologists also predicted whether or not repeat images would be requested by the radiologist. R~ul ts : Three-hundred eighty CT exams were studied [outpatients (277), inpatients (90), El) patients (9), ICU patients (4)]. Mean total exam time was 44.7 minutes (image time 33.1 minutes; review time 8.2 mim~es; and dear time 3.4 mitmtes). Conmltation with a senior radiologist was requested statistically significantly more frequently: (a)by junior re, sidents than by senior residents or fellows, and (b)in ED and ICU patients (22% and 50%, rospe~vely) than in outpatients and inpatients (10% and 14%, respoctively). Repeat images were obtained in 75 patients not significantly related to patient status, scan type, or radiologist training level. W h ~ the teelmologist predicted that no repeat images were needed, this prediction was correct in 86% of the cases. ~ : Reviewing scans before the patient leaves the CT suite adds considerably to the total time required to complete a scan. If technologists obtained repeat images at their discretion, time would be saved.

10-7 MEDICAL ERRORS IN RADIOLOGY: OBSERVATIONS AND LESSONS FROM A PROSPECTIVE VOLUNTARY "TREASURE" HUNT, Christos A. Athanasoulis M.D., Massachusetts General Hospital, Harvard Medical School, Boston, MA ~ : Medical errors may be looked upon not as human failures but rather as "treasures" i.e. opportunities for improvement. In this context, a"treasurelaunt" is an effort to discover and to collect errors. Our aim was to test the hypothesis that members of the radiology staff would be willing to participate in a prospective "treasure hunt" and that voluntary retxming of errors might be feasible. An additional aim was to evaluate the nature of recorded errors. ~ : Medical error was defined as an unintended act (either of omission or commission) or an act that did not achieve its intended outcome. After appropriate introduction the radiographers, nurses and radiologists of one department unit were poUed regarding their interest in learning from errors. Those who were interested were prompted to record errors and to submit them anonymously for tabulation. Errors were classified as medical, administrative and teehni.caI or equipment related. Results: During the initial poll eighteen of twenty staff members (90%) voted in favor of prospectively recording errors. Two voted against. During a three month period, sixty events or errors were recorded. Forty one errors (68%) were recorded during the fwst month (the period of enthusiasm), 27% during the second month and a mere 5% in the final month. 70% were errors of omission while 30% were errors of commission. Thirty one errors (52%) were of administrative nature, twenty (33%) were of medical nature and nine (I5%) were technical or equipment related. No error resulted in major physical harm to the patient. n.Cr~l.u~9_~: Medical errors of varied nature did occur. None resulted in serious harm to the patient. A voluntary anonymous system for the ongoing identification and recording of errors did not succeed due to loss of interest after an initial burst of enthusiasm.