Chapter 6. Reengineering

1. Chapter 6.Reengineering Yasar A. Ozcan

2. Outline • Reengineering vs. Other Methods • Work Design in Health Care Organizations • Work Design • Job Design • Work Measurement-Standard Times • Stopwatch Time Studies • Standard and Predetermined Times • Work Measurement Using Work Sampling • Determination of Sample Size • Development of Random Observations Schedule • Training Observers • Work Simplification • Flow Chart • Work Distribution Chart • Flow Process Chart • Worker Compensation Yasar A. Ozcan

3. Reengineering vs. Other Methods • Healthcare managers have often sought organizational change, restructuring, and downsizing. Although those methods may improve the financial base of the organization or productivity at least temporarily by “cutting the fat,” namely by reducing the staff across the board, yet they create other problems. In particular, reducing staff can lead to major problems in the quality of care. • Two other contemporary and popular methods that aim to improve both performance and the quality are total quality management (TQM) and continuous quality improvement (CQI) which are geared to make incremental changes over time. Yasar A. Ozcan

4. Reengineering vs. Other Methods • Reengineering is a methodology intended to overcome the difficulty in realizing TQM/CQI performance over a long duration, as well as the myopic conduct of organizational change, restructuring and downsizing. • To reengineer the system, healthcare managers must be able to understand work-design, jobs, job measurement, process activities, and reward systems – all well known concepts of industrial engineering. With that knowledge, they can recognize the bottlenecks in the old system, identify unnecessary and repetitive tasks, and eliminate them in the reengineered system of care. Yasar A. Ozcan

5. Importance and Purpose • Human resources represents over 40% of healthcare facility budgets • Productivity and satisfaction of staff involves an understanding of the work environment • Work must be designed so that employees are happy, organizational productivity is high, and costs are minimized Yasar A. Ozcan

6. Work Design • Work • Measurement • Time Study • Predetermined • Standard • Work Sampling • Job Design • Who? • How? • Where? • Job Simplification Work Design- A Systems Perspective External Factors • Worker • Compensation • Time Based • Output Based • Incentive Plans Yasar A. Ozcan

7. Job DesignWho does what, how, and where? • Consistent with organizational goals • Write it down! • Understand and communicate it • Involve employees and management Yasar A. Ozcan

8. Frederick Winslow Taylor • Developed Scientific Management Approach • Focused on time studies • Conflicts between labor and management occurred because management had no idea how long jobs actually took For what types of jobs would this approach work best? Are there Healthcare applications? Yasar A. Ozcan

9. Efficiency School--Logical and Systematic • Best for simple, repetitive routine, and separable tasks • Healthcare Examples: • lower level administrative duties • division of labor • standardized forms and paperwork • robots in laboratories • automation of routine tasks • Not good for judgmental/unpredictable tasks Yasar A. Ozcan

10. Advantages and Disadvantages of Specialization Yasar A. Ozcan

11. Behavioral SchoolSatisfaction of Wants/Needs • Intrinsic and extrinsic motivators • Specialization leads to monotony and worthlessness Yasar A. Ozcan

12. Behavioral School (Human Focus) Socio-technical School Socio-technical School Approach Efficiency School (Technical Focus) Yasar A. Ozcan

13. How can jobs be improved? • Job enlargement-- give workers a larger portion of the total task (horizontal loading-- additional work at same level of skill and responsibility) • Job enrichment-- increasing responsibility for planning and coordinating tasks (vertical loading) • Job rotation-- workers periodically exchange jobs What are examples of each? Yasar A. Ozcan

14. Work Measurement Using Time Standards • Time standards are important in establishing productivity measures, determining staffing level and schedules, estimating labor costs, budgeting, and designing incentive systems • A time standard represents the amount of time needed for the average worker to do a specific job working under typical conditions Yasar A. Ozcan

15. The First Step. . . The amount of time it should take a qualified worker to complete a specified task, working at a sustainable rate, using given methods and equipment, raw materials, and workplace arrangements is called a standard time. • A Standard timecan be developed through: • Stop-watch studies • Historical times • Predetermined data • Work sampling Yasar A. Ozcan

16. Stopwatch Time Studies • Take time over a number of trials (cycles) • Workers should be educated regarding the process to avoid suspicion and avoid the Hawthorn Effect • Number of cycles to time (i.e., sample size) • variability in observed times • desired accuracy • desired level of confidence for the estimate Yasar A. Ozcan

17. 2 ( z s a x ) n = Determining Sample Size Accuracy desired may be explained by the percentage of the mean of the observed time. For instance, the goal may be to achieve an estimate within 10 percent of the actual mean. The sample size is then determined by: where: z = number of std. dev. needed for desired confidence s = sample std. dev. a = desired accuracy x = sample mean Desired Confidence Z-value 90 1.65 95 1.96 98 2.33 99 2.58 Yasar A. Ozcan

18. 2 ( zs e ) n = An Alternative Formula Desired accuracy may be expressed as an amount (e.g., within one minute of the true mean). The formula for sample size becomes: where e = Accuracy or maximum error acceptable To make an initial estimate of sample size, you should take a small number of observations and then compute the mean and std. dev. to use in the formula for n. Yasar A. Ozcan

19. 2 ( zs ax ) 2 ( zs e ) n = n = An Example A time study analyst wishes to estimate the time required to perform a certain job. A preliminary study yielded a mean of 6.4 minutes and a standard deviation of 2.1 min. The desired confidence level is 95 percent. How many observations will be needed (including those already taken) if the desired maximum error is: a) +/- 10 percent? b) one-half minute? a) b) 2 2 ( 1.96(2.1) .10(6.4) ) ( 1.96(2.1) 0.5 ) = 41.36 = 67.77 = n = Yasar A. Ozcan

20. Determining the Standard Time • Observed Time-- average of observed times • OT = åxi/n • Normal Time-- observed time adjusted for worker performance • NT = OT * PR (where PR = performance standard measured for the entire job) • NT = å(Ej*PRj) (where PR is measured element by element) • PR equals 1 for the average worker; PR< 1 is for a slower worker Yasar A. Ozcan

21. ST = NT * AF • Standard time equals normal time multiplied by an allowance factor • Allowance Factor • accounts for personal delays, unavoidable delays, and/or rest breaks • AFjob = 1+A, where A= allowance percentage based on job time • AFday = 1/(1-A), where A = allowance percentage based on work day Yasar A. Ozcan

22. Table 6.1 Typical Allowance Percentages for Varying Healthcare Delivery Working Conditions Source: Adapted from B.W. Niebel, 1988. Yasar A. Ozcan

23. The Allowance Factor • Compute the allowance factor if: • The allowance is 20 percent of job time. • The allowance is 20 percent of work time. A) AF = 1 + A = 1.20, or 120% B) AF = 1/(1-A) = 1/(1-.2) = 1.25 = 125% Yasar A. Ozcan

24. Obs Time (min) Obs Time 1 4.20 6 4.18 2 4.15 7 4.14 3 4.08 8 4.14 4 4.12 9 4.19 5 4.15 37.35 An Example The following observed times have been collected with a performance rating of 1.10. Using an allowance factor of 15% of job time, determine the appropriate standard time. Preliminary Calculations: n=9 PR =1.10 A = 1.15 Yasar A. Ozcan

25. Obs. Time (min) Obs. Time 1 4.20 6 4.18 2 4.15 7 4.14 3 4.08 8 4.14 4 4.12 9 4.19 5 4.15 37.35 The Solution A. OT = 37.35/9 = 4.15 minutes B. NT = OT x PR = 4.15 x 1.10 = 4.565 minutes C. ST = NT x (1+A) = 4.565 x 1.15 = 5.25 minutes Now, what do: OT, NT and ST represent? Yasar A. Ozcan

26. What are the problems with time studies? • Subjective performance ratings and allowances • Only observable jobs can be studied • Highly costly -- best for repetitive tasks • Disrupts worker routine • May cause worker resentment Yasar A. Ozcan

27. Other Methods • Historical/Standard Elemental Times • Firms collect data on standard job elements • Put these data together to determine job times • Less costly and disruptive • Limited applications in healthcare • Predetermined Standards • Obtained from trade publications • Need no performance of allowance factor • Operations are not interrupted Yasar A. Ozcan

28. Work Measurement Using Work Sampling • Technique for estimating the proportion of time that a worker or machine spends on various activities • Observers make brief observations of a worker or a machine at random intervals over a period of time and simply note the nature of the activity • Purpose: • To estimate percentage of unproductive or idle time for repetitive jobs • To estimate the percentage of time spent on various tasks for non-repetitive jobs Yasar A. Ozcan

29. Work Sampling Steps 1) Determine the sample size 2) Train the observers, 3) Develop random sample schedule 4) Take observations, and re-compute the desired sample size several times if initial estimates are not reliable 5) Determine the estimated proportion of time spent on specified activity Yasar A. Ozcan

30. Step 1: Sample Size The goal of work sampling is to obtain an estimate that provides a specified confidence not differing from the true value by more than a specified error CI = confidence interval, e = error, z = number of standard deviations needed to achieve desired confidence, sample proportion (number of occurrences divided by sample size), n = sample size. Yasar A. Ozcan

31. Sample Size, cont. Example 6.1:A hospital administrator wants an estimate of X-ray idle time that has a 95.5 percent confidence of being within 4 percent of the actual percentage. What sample size should be used? e = 0.04 z = 2.00 n = (z/e)2p(1-p) Desired Confidence Z-value 90 1.65 95 1.96 95.5 2.00 98 2.33 99 2.58 When p is unknown, a preliminary estimate of sample size can be obtained using p = 0.5. Then after 20 observations, a new estimate can be obtained. Yasar A. Ozcan

32. Solution: Given: e = 0.04; z = 2.00 (see Appendix A); = 0.5 (preliminary). = 0.5: n = (2.00/0.04)2 * .50 * (1-.50) = 625 observations. If for 20 observations, it is observed that the x-ray was breaking down on average 1 time, the revised estimate is then = 1/20 = 0.05. The revised estimate of sample size is: = 0.05, n = (2.00/0.04)2 x .05 x (1-..05) = 118.75 or 119 observations. Yasar A. Ozcan

33. Step 2: Train Observers A comprehensive training program of three steps should be standardized for all data collectors. Data collectors should be first educated as to the study’s goals, protocol, collection procedures, and data submission procedures, and the guidelines for their behavior. Then, the observers should be trained in data collection. Training may include sessions using videotaped activities for practice in identifying and recording actual nursing services. In the third phase, observers participate with a project member, in explaining the nature of the project to those who will be observed, in the observation setting. Yasar A. Ozcan

34. Step 3: Random Observation Schedule • Need random number for day, hour, and minute, with the number of digits needed for each number equaling the number of days in the study, hours per day, and minutes per hour. Yasar A. Ozcan

35. Table 6.6 Random Numbers Yasar A. Ozcan

36. Example 6.3:A nursing manager wants to observe the time a nurse spends in direct and in indirect care over a 5-day period, on a unit where the shift is 8 hours. Solution: • A one digit number will be needed for the day, one digit for the hour, and two digits for the minute. • Using Table 6.4, starting from row 4, column 3, we obtain the random number 59830829. • The first number is 5. Thus, we determine the day (in this case, the fifth day of the week, Friday). • We move to the next number, 9, for the hour; but since activity is performed 8 hours daily, we discard that number, and move to the next one, 8. If we assume that the shift starts at 7:00 AM, the number 8 represents 2:00 PM. • The minutes are derived from the next two digits, 30. Put together, the first observation is made on the fifth day, 30 minutes into the 8th hour of work, or at 2:30 PM. • This procedure is repeated for each observation to be taken. Then the observations should be sorted chronologically by day, hour and minute. Yasar A. Ozcan

37. Example 6.4:The chief of the hospital maintenance technicians wants to estimate the proportion of time that technicians spend in a part of the maintenance process. The maintenance office is open 9 hours, starting at 8:00 AM, every day of the week. 20 observations will be taken during a month-long investigation. Determine the random observation times and develop an observation schedule, assuming that the serial number of a dollar bill starts with 25. Solution: • Since we know the starting point in Table 6.4 is the 2nd row and 5th column, the random number is 53664329. • Next we need to choose the reading direction of the succeeding random numbers: either by moving to the right on the same row, and when the row is finished, going down one row and moving from left to right; or by going down on the same column, and when the column is finished moving to the next column right and reading from bottom to top. • For this case, we choose to read in the same column, going down. For days, read two digits from left to right – select two-digit number, if higher than 31, then move to the next digit to make a two-digit day observation, and so on. • Within eight digit numbers, if there are not enough digits to identify day, hour and minutes for the observation, discard that random number and select the next one. • For hours, read single digit numbers from left to right, discard 0 and assign 1=8:00 A.M., 2=9:00 A.M., and so on. • For minutes, read two digit numbers and discard numbers 60 or over. Prepare a chronological list of the observation time results by day, hour and minute, to be given to the data collection team. Yasar A. Ozcan

38. Table 6.7 Development of the Schedule for a Work Sampling Study Yasar A. Ozcan

39. Table 6.8 Final Work Sampling Schedule Yasar A. Ozcan

40. Table 6.4 Abridged Patient Care Tasks in a Nursing Unit Yasar A. Ozcan

41. Table 6.5 Work Sampling Data Collection Form for Nursing Unit Yasar A. Ozcan

42. Work Sampling Steps 4) Take observations, and re-compute the desired sample size several times if initial estimates are not reliable 5) Determine the estimated proportion of time spent on specified activity Yasar A. Ozcan

43. Advantages of Work Sampling • Observations less susceptible to short term fluctuations • Little or no work disruption • Workers are less resentful • Less costly and time-consuming • Many studies can be conducted simultaneously • Useful for non-repetitive tasks Yasar A. Ozcan

44. Disadvantages of Work Sampling • Less detail on elements/tasks of a job • Workers may alter patterns • Often no record of method used by worker • Observers may fail to adhere to random observation schedule • Not useful for short, repetitive tasks • Much time required to move from observation area to observation area to ensure randomness Yasar A. Ozcan

45. Nobody likes to do things the hard way!Work Simplification • Work Simplification-- process of changing work methods: • Eliminate unnecessary parts of work • Combine and rearrange parts of work • Simplify work when possible • Work Simplification Tools • Flow Chart • Work Distribution Chart • Flow Process Chart • Layout Chart Yasar A. Ozcan

46. Figure 6.4 Commonly Used Flow Chart Symbols Start/Terminate On page connector Process Off page connector Decision Document Preparation Manual Operation Yasar A. Ozcan

47. Figure 6.5 Flow Chart for Emergency Room Specimen Processing Initial Process After Improvement Patient Entry Patient Entry Triage: need blood? Triage: need blood? No end No end Yes Nurse draws blood Nurse draws blood MD orders lab MD orders lab IS order entry IS entry label & package Lab Label & package Lab Results arrive in ER (end) Accession & analysis Verification Accession & analysis IS double entry MD terminates lab order (end) Yasar A. Ozcan

48. The Work Distribution Chart • Shows what a department does to identify each of its major activities and to pinpoint the contribution of each employee to those activities • Must be specific! • Spotting Trouble • Which activities consume the most time? • Are tasks evenly distributed? • Is there under-specialization? • Are employees assigned too many unrelated tasks? • Are talents utilized efficiently? Yasar A. Ozcan

49. Table 6.9 Partial Work Distribution Chart for Nursing Unit Yasar A. Ozcan

50. Flow Process Chart • Records a procedure in a graphic form, using a sort of shorthand to simplify and unify the record • Ensures every significant detail of the work process in its proper sequence is recorded • Highlights inconsistencies and redundancies • Can eliminate, combine, change (sequence, place, person), or improve activities Operation Inspect Store Move Delay Yasar A. Ozcan