Assessing the precision of a time-sampling-based study among GPs: balancing sample size and measurement frequency

Table 4 Total CIs based on simulations conducted on a 40-h working week and differences with the total CIs based on simulations conducted on six alternative working schedules, divided by the frequency of the measurements during 1 week, including 50 and 100 GPs

	Measurement per:
	0.5 h	1 h	2 h	3 h
Total CI based on 40-h working schedule used (50 GPs)	1.17	1.71	2.44	3.09
Change of total CI in case of
• Alternative 1	− 0.04	− 0.07	− 0.12	− 0.21
• Alternative 2	− 0.04	− 0.05	− 0.09	− 0.14
• Alternative 3	− 0.03	− 0.07	− 0.10	− 0.13
• Alternative 4	− 0.03	− 0.06	− 0.08	− 0.08
• Alternative 5	− 0.03	0.02	0.10	0.06
• Alternative 6	− 0.02	0.05	0.07	0.08
Total CI based on 40-h working schedule used (100 GPs)	0.83	1.21	1.73	2.15
Change of total CI in case of
• Alternative 1	− 0.03	− 0.05	− 0.08	− 0.12
• Alternative 2	− 0.03	− 0.03	− 0.05	− 0.07
• Alternative 3	− 0.02	− 0.05	− 0.07	− 0.06
• Alternative 4	− 0.02	− 0.04	− 0.06	− 0.02
• Alternative 5	− 0.02	0.02	0.08	0.05
• Alternative 6	− 0.02	0.03	0.05	0.10

The total CI is based on the combination of sample and measurement fluctuation as has been explained in the “Methods” section. Simulations on a 40-h working schedule are conducted to assess the amount of measurement fluctuation. All working days start at 7:30 am. The beginning of the working day by a half an hour time slot was set at a couple of minutes from 7:30 am to simulate the CIs
Legend: Working schedule used in this paper = 5 working days of 8 h; 1-h break (40 h per week). Alternative 1 = 2 working days of 8 h; 1-h break (16 h per week). Alternative 2 = 3 working days of 8 h; 1-h break (24 h per week). Alternative 3 = 4 working days of 4 h; no break (16 h per week). Alternative 4 = 5 working days of 4 h; no break (20 h per week). Alternative 5 = 6 working days of 10 h; 1-h break (60 h per week). Alternative 6 = 7 working days of 8 h; 1-h break (56 h per week)

ISSN: 1478-4491