Optimised Parameters of Pharmacy Robotic System (PRS) For Optimised Parameters of Pharmacy Robotic System (PRS) For Hospital and Polyclinic PharmaciesHospital and Polyclinic Pharmacies

Siang Li CHUA Health Services & Outcomes Research, National Healthcare Group

Background

A Pharmacy Robotic System (PRS), to replace manual drug picking and packing at a hospital outpatient pharmacy, was being designed to improve patient safety. As the robot can be a common system for the institutions and polyclinics, purchasing a few systems will be more cost effective provided that the designed robot can meet the requirements of all sites.

Objective

Different specifications of a robotic system yield distinict outcomes. Simulation models are cost-effective and flexible tools that is able to provide such pre-designed specifications of a new system.

The objective of this study is to determine the speed of a common PRS so as to achieve Wait Time (WT) target of each of the four hospitals and two polyclinic pharmacies.

Data

For each site, one or two months of historical data (between Jan to Mar 06) of patients’ ticket issue times, pharmacy WT and turnaround time (TAT) from the system, staff schedule and process service times were collected and were analysed to quantify:

a) Demand: Total number of drugs dispensed daily (Figure 1) and the number of items per prescription (Figure 2).

b)Rate of demand: Half-hourly prescription arrival rate at the pharmacy counter (Figure 3).

c) Supply: Number of each staff type on-duty every half-an-hour. The staff type are prescription transcribers, packer, checker and dispenser.

d)Rate of dispensing and billing: Time taken to transcribe prescriptions into the system, packing, checking and dispensing of drugs, and paying bills. These times were manually recorded and noted by the staff.

Results

Data analysis showed that the average number of items dispensed daily at the 6 pharmacies ranged from 681 to 2697. At four sites, 27% to 38% of their prescriptions are 1-item prescription while at the other two sites, they have 2-item prescription at 23% to 27%. The average number of daily prescriptions ranged from 222 to 960. Site 1has the most number of items dispensed and prescriptions daily.

The simulation model showed that Site 1 needs the fastest machine and it determine the overall speed design of the robot. See Table 1.

To implement a common PRS system across the four hospitals and 2 polyclinics, PRS packing speed has to be 9 seconds per item so that all pharmacies can meet the targeted WT.

Conclusion

The work group is sourcing for a 9-seconds per item robot.

Contact details: Chua Siang Li, siang_li_chua@nhg.com.sg, Office: 64966930/63892185

Method

Twelve simulation models, two at each site, were built to study the existing process and process with PRS. In the existing system, once the patient obtained the queue number, the prescription would be manually transcribed into the system. The drugs would be manually packed, checked and dispensed. With PRS, each item would be machined packed and consolidated by prescription. Each packed prescription is then manually assembled, checked and dispensed. WT is from the time the patient arrives till the time he is called for dispensed.

Figure 4 and Figure 5 show the simulation model of current process and process with PRS of Site 1 respectively.

Figure 4. Simulation model of current manual process (Site 1)

Figure 5. Simulation model of process with PRS (Site 1)

Table 1. Simulation results (in last row)

Figure 1. Daily average total items dispensed

Figure 3. Daily average number of patients arriving at the pharmacy every half-an-hour on weekday

0

10

20

30

40

50

60

70

80

08

00

- 0

83

00

83

0 -

09

00

09

00

- 0

93

00

93

0 -

10

00

10

00

- 1

03

01

03

0 -

11

00

11

00

- 1

13

01

13

0 -

12

00

12

00

- 1

23

01

23

0 -

13

00

13

00

- 1

33

01

33

0 -

14

00

14

00

- 1

43

01

43

0 -

15

00

15

00

- 1

53

01

53

0 -

16

00

16

00

- 1

63

01

63

0 -

17

00

17

00

- 1

73

01

73

0 -

18

00

18

00

- 1

83

01

83

0 -

19

00

Site 1 (Mar 06)

Site 2 (Feb 06)

Site 3 (Feb-Mar 06)

Site 4 (Jan 06)

Site 5 (Mar 06)

Site 6 (Mar 06)

Figure 2. Percentage on Number of Items in a Prescription

1

11

1

22

0%

5%

10%

15%

20%

25%

30%

35%

40%

Site 1 Site 2 Site 3 Site 4 Site 5 Site 6

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

Site Site 1 Site 2 Site 3 Site 4 Site 5 Site 6Daily average total items dispensed

2697 2244 2390 1992 1129 681

Most frequent number of items per prescription and its percentage

1 (37.6%)

2 (25.6%)

2 (22.9%)

1 (33.0%)

1 (32.3%)

1 (26.7%)

Daily average number of prescription

 960 (Mar 06)

713 (Feb 06)

704 (Feb-

Mar 06)

600 (Jan 06)

402 (Mar 06)

222 (Mar 06)

Manual process

Transcribe prescription

man pack sort check dispense

Transcribe prescription man pack check dispense

Man pack check dispense

Process with PRS

Transcribe Prescription PRS pack check dispense

PRS pack

check dispense

WT target 30 mins for 95th percentile20 min

max

20 min 80th %tile

Desired PRS service time/ item

9 seconds

12 seconds

 11 seconds

14 seconds

16 seconds

25 seconds

-

500

1,000

1,500

2,000

2,500

3,000

Total Items 2,697 2,244 2,390 1,992 1,129 681

Site 1 Site 2 Site 3 Site 4 Site 5 Site 6