Table 2.

Communication Pattern Results from Behavior Mapping Observation and Social Network Survey

Approach	Clinic A	Clinic S
Behavior mapping observation
Observation rounds (interval between rounds)	37 (15 min)	26 (30 min)
Data points*	667	1346
Data length	9.25 h	13 h
Social network survey†
N (response rate, total staff members)	6 (19%, 32)	18 (23%, 78)
Face-to-face “routine work” communication
Data points‡	121	717
ANOVA test for mean distances per communication frequency significance§	p < 0.001‖	p < 0.001‖
Virtual “routine work” communication
Data points‡	101	611
ANOVA test for mean distances per communication frequency significance§	p = 0.23	p < 0.001‖
Face-to-face “seeking advice” communication
Data points‡	40	214
ANOVA test for mean distances per communication frequency significance§	p = 0.04	p < 0.001
Virtual “seeking advice” communication
Data points‡	31	177
ANOVA test for mean distances per communication frequency significance§	p = 0.004‖	p < 0.001‖

* Each data point represents a single individual

† The social network survey responses were translated into communication relationships for each type of communication. The low response rate for the personal interest communication type provided too few data points to include in the analysis

‡ Each data point represents communication frequency between two individuals

^§ Mean distances between staff members per communication frequency and post hoc Tukey HSD multiple comparison results are illustrated in Fig. 4

^‖ An adjusted F test was performed using Brown-Forsythe statistic since an assumption of homogeneity of variance was not met