From Jul 2017 to Mar 2020, 637 tracing reports were submitted by the Nihon Chouzai Koiwa-Minamiguchi Pharmacy to different medical institutions, including hospitals and clinics. Of these, 177 included deprescribing proposals. Patients consented to the submission of the reports to prescribing doctors in 149 of 177 cases. Cases in which multiple reports were submitted to the prescribing doctor, a single report was included in the analysis. Finally, 123 tracing reports that included deprescribing proposals were analyzed (Fig. 1). Deprescription succeeded in 67 cases and failed in 56 cases. The deprescription success rate was 54.5%
In the univariate analysis, the P-values of sex, potential inappropriate prescriptions, family pharmacists, unused medications, possible adverse events, and overlapping same effective medications were less than 0.2. Therefore, these endpoints were analyzed using multivariate logistic regression analysis (Table 1). Comparing the successful deprescribing group vs the unsuccessful deprescribing group, covariates of average age (76.0 years old vs 76.2 years old, odds ratio: 0.999), average number of oral medications prescribed (8.2 medications vs 7.8 medications, odds ratio: 1.03), average dosing frequency per day (3.6 times vs 3.3 times, odds ratio: 1.14), % of one-dose packages (32.8% vs 26.8%, odds ratio: 1.34), and % of “pharmacist experienced face-to-face communication” (76.1% vs 66.1%, odds ratio: 1.64) indicated a P-value > 0.2 and were not included in the multivariate analysis. Meanwhile, sex (% of male) (35.8% vs 51.8%, odds ratio: 0.52), % of potential inappropriate prescriptions (64.2% vs 42.9%, odds ratio: 2.39), % of family pharmacists (70.1% vs 53.6%, odds ratio: 2.04), % of unused medications (61.2% vs 37.5%, odds ratio: 2.63), % of possible adverse events (73.1% vs 44.6%, odds ratio: 3.38), and % of overlapping same-effect medications (64.2% vs 42.9%, odds ratio: 2.39) indicated a P-value < 0.2 and were included in the multivariate analysis.
Table 1
Results of univariate analysis performed for each endpoint
|
Deprescribing succeeded
(n = 67)
|
Deprescribing failed
(n = 56)
|
Odds ratio
|
P-value of univariate analysis
|
Average age
|
76.0
|
76.2
|
0.999
|
0.966
|
Average number of oral medications prescribed
|
8.2
|
7.8
|
1.03
|
0.51
|
Average dosing frequency per day
|
3.6
|
3.3
|
1.14
|
0.32
|
Sex (% of male, female)
|
(35.8, 64.2)
|
(51.8, 48.2)
|
0.52
|
0.0764
|
% of potential inappropriate medications
|
64.2
|
42.9
|
2.39
|
0.019
|
% of family pharmacists
|
70.1
|
53.6
|
2.04
|
0.06
|
% of one-dose packages
|
32.8
|
26.8
|
1.34
|
0.467
|
% of unused medications
|
61.2
|
37.5
|
2.63
|
0.00956
|
% of possible adverse events
|
73.1
|
44.6
|
3.38
|
0.00158
|
% of overlapping same-effect medications
|
64.2
|
42.9
|
2.39
|
0.019
|
% of “pharmacist experienced face-to-face communication”
|
76.1
|
66.1
|
1.64
|
0.221
|
Univariate logistic regression analysis was performed. Covariates with a P-value of less than 0.20 in the univariate analyses were used in multivariate logistic regression analysis. |
In the multivariate analysis, the presence of unused medications (odds ratio: 3.21, 95% CI: 1.4–7.38, P-value: 0.00593) significantly increased the deprescribing success rate (Table 2, Fig. 2). In addition, sex (male) (odds ratio: 0.547, 95% CI: 0.242–1.24, P-value: 0.147), potential inappropriate medications (odds ratio: 1.75, 95% CI: 0.738–4.13, P-value: 0.205), overlapping same-effect medications (odds ratio: 1.91, 95% CI: 0.839–4.37, P-value: 0.123), and possible adverse events (odds ratio: 2.55, 95% CI: 0.993–6.54, P-value: 0.0518) tended to increase the deprescribing success rate but not significantly.
Table 2
Results of logistic regression analysis
|
Odds ratio
|
95% CI
|
P-value of logistic regression analysis
|
Family pharmacists
|
1.08
|
0.458
|
2.53
|
0.867
|
Unused medications
|
3.21
|
1.4
|
7.38
|
0.00593**
|
Sex (male)
|
0.547
|
0.242
|
1.24
|
0.147
|
Potential inappropriate medications
|
1.75
|
0.738
|
4.13
|
0.205
|
Overlapping same-effect medications
|
1.91
|
0.839
|
4.37
|
0.123
|
Possible adverse events
|
2.55
|
0.993
|
6.54
|
0.0518
|
Results of multivariate logistic regression analysis. * indicate P-value < 0.05, ** indicate P-value < 0.01. 95% CI indicate 95% confidence interval. |