Of the 8278 hospitals that were part of the annual survey, responses were received from 3612 hospitals (recovery rate: 43.6%). Out of these 3612 hospitals, 210 hospitals meeting the criteria for ARIP 1 with 100–299 beds, and 245 hospitals meeting the criteria for ARIP 2 with 100–299 beds, were included in our analysis (Fig. 1). Among the pharmacist services, significant differences were observed between ARIP 1 and 2 hospitals in “sterile dispensing services for inpatients,” “sterile dispensing services for outpatients,” and “therapeutic drug monitoring (TDM) services.” ARIP 1 hospitals were more frequently graded as “Fairly well (more than 80%)” and “Often (more than 50%)” than ARIP 2 hospitals. ARIP 1 hospitals were more frequently assessed as “Fairly well” or “Often” for working on education/research (including trainee guidance) (Table 2).

Extraction criteria and flow diagram of selected hospitals for analysis

To understand the characteristics of the subject facility, multiple linear regression analysis was performed. The objective variable was the average length of stay Median value, and the explanatory variables were ARIP category, DPC classification, number of beds, and number of pharmacists. DPC classification (B = − 28.234, β = − 0.160, P = 0.003, VIF = 1.417), number of beds (B = 0.321, β = 0.197, P = 0.001, VIF = 1.730), and number of pharmacists (B = − 2.734, β = 0.161, P = 0.007, VIF = 1.683) were extracted as factors affecting the average length of hospital stay. Thus, to better align backgrounds, a comparison of hospitals subject to DPC for ARIP 1 and 2 (ARIP 1: 173 hospitals; ARIP 2: 105 hospitals) was conducted. The results revealed a significant difference in the number of pharmacists, with a larger number in ARIP 1 hospitals. Among the pharmacist services, significant differences were observed in “Pharmaceutical management and guidance to pre-hospitalization patients,” “Dispensing and management guidance for outpatients,” “sterile preparation and processing services for inpatients,” and “sterile preparation and processing services for outpatients,” with a larger number in ARIP 1 hospitals. ARIP 1 hospitals generally more participation in educational and research activities. No difference was observed in the rate of issuance of outpatient prescriptions between ARIP 1 and 2 hospitals (Table 3). Additionally, a weak negative correlation was observed between the number of pharmacist services with “Fairly well” or “Often” and the number of beds per pharmacist for both ARIP 1 (R = -0.207) and ARIP 2 (R = -0.279) DPC hospitals (Fig. 2).

Relationship between state of implementation of pharmacist services and number of beds per pharmacist (DPC hospitals). a Additional reimbursement for infection prevention category 1 and DPC hospitals. b Additional reimbursement for infection prevention category 2 and DPC hospitals

In DPC hospitals with ARIP 1 and 2, the average number of beds per pharmacist (95% CI) was 21.7 (20.4–23.1) and 24.7 (22.6–26.8) (P < 0.05), and the average length of stay (d) (median value: min–max) was 14.3 (12.3–17.0) and 15.4 (13.7–24.0) (P < 0.01), respectively (Table 3). No correlation between the average number of beds per pharmacist and the average length of hospital stay was observed for ARIP 1, but however, a weak correlation (R = 0.322) was observed for ARIP 2 (Fig. 3). To confirm this result and identify the effect of other factors, multiple linear regression analysis was performed. As a result, though no factors affecting the average length of stay were extracted in ARIP 1 hospitals (number of beds [B = 0.001, β = 0.07, P = 0.950, VIF = 1.991], number of pharmacists [B = − 0.158, β = − 0.161, P = 0.307, VIF = 4.334], and beds per pharmacist [B = 0.013, β = 0.023, P = 0.875, VIF = 3.665]), beds per pharmacist was extracted as an affecting factor on the average length of stay was extracted in ARIP 2 hospitals (number of beds [B = − 0.033, β = − 0.237, P = 0.092, VIF = 2.254], number of pharmacists [B = 0.383, β = 0.282, P = 0.173, VIF = 4.881], and beds per pharmacist [B = 0.284, β = 0.521, P = 0.003, VIF = 3.512]).

Relationship between the average length of hospital stay and number of beds per pharmacist (DPC hospitals). a Additional reimbursement for infection prevention category 1 and DPC hospitals. b Additional reimbursement for infection prevention category 2 and DPC hospitals