Exploring thresholds and interaction effects among antibiotic usage, covariates, and their effect on antibiotic resistance using an extended-spectrum β-Lactamase–producing Klebsiella pneumoniae case

Figures & data

Figure 1. Monthly ESBL-producing K. pneumoniae incidence versus use of third-generation cephalosporin (5-month moving averages).

Figure 2. Monthly ESBL-producing K. pneumoniae incidence versus use of alcohol-based hand rub (5-month moving averages).

Table 1. Linear regression modeling results for ESBL-producing Klebsiella pneumoniae, January 2017 to December 2021; adjusted R-squared = 20.03%.

Predictor Variable	Lag	Median use (IQR)	Coefficient (95% CI)	p-value
Constant	NA	NA	0.0589 (0.0415 to 0.0762)	<0.001
Third-generation cephalosporin use (DDD/100 OBD)	1	2.33 (1.77–3.27)	0.0129 (0.0059 to 0.0198)	<0.001
Alcohol-based hand rub (L/100 OBD)	1	4.89 (3.97–7.54)	−0.0033 (−0.0051 to −0.0015)	<0.001

Table 2. MARS (Additive) result in modeling ESBL-producing Klebsiella pneumoniae incidence, January 2017 to December 2021; adjusted R-squared = 22.47%.

Predictor Variable	Lag	Median use (IQR)	Threshold (confidence limit)*	Relation to threshold	Cases satisfying threshold condition	Coefficient (95% CI)	p-value
Constant	NA	NA	NA	NA		0.0689 (0.0603 to 0.0774)	<0.001
Third-generation cephalosporin use (DDD/100 OBD)	1	2.33 (1.77–3.27)	2.40 (1.62–3.83)	Above	48%	0.0147 (0.0069 to 0.0225)	<0.001
Alcohol-based hand rub (L/100 OBD)	1	4.89 (3.97–7.54)	6.51 (3.62–9.60)	Above	28%	−0.0037 (−0.0057 to −0.0018)	<0.001

Note: * 5^th and 95^th percentiles of significant thresholds in OAT sensitivity analysis used as limits around threshold; NA, not applicable.

Table 3. MARS (Interaction, i = 2) result in modeling ESBL-producing Klebsiella pneumoniae incidence, January 2017 to December 2021; adjusted R-squared = 29.77%.

Predictor Variable	Lag	Median use (IQR)	Threshold (confidence limit)*	Relation to threshold	Cases satisfying threshold condition	Coefficient (95% CI)	p-value
Constant	NA	NA	NA	NA	NA	0.0765 (0.0674 to 0.0856)	<0.001
Third-generation cephalosporin use (DDD/100 OBD)	2	2.33 (1.77–3.27)	2.00 (1.62–3.83)	Above	59%	0.0109 (0.0041 to 0.0177)	0.0026
Alcohol-based hand rub (L/100 OBD)	1	4.89 (3.97–7.54)	6.61 (3.62–9.60)	Above	28%	−0.0083 (−0.0121 to −0.0046)	<0.001
Third-generation cephalosporin use (DDD/100 OBD)	1	NA	3.71 (1.62–3.83)	Above	19%	0.0045 (0.0017 to 0.0072	0.0027
Alcohol-based hand rub (L/100 OBD)	1	NA	6.61 (3.62–9.60)	Above
Third-generation cephalosporin use (DDD/100 OBD)	1	NA	3.13 (1.62–3.83)	Below	33%	0.0026 (0.0005 to 0.0048	0.0174
Third-generation cephalosporin use (DDD/100 OBD)	2	NA	2.00 (1.62–3.83)	Above

Note: * 5^th and 95^th percentiles of significant thresholds in OAT sensitivity analysis used as limits around threshold; NA, not applicable.

Table 4. Absolute and relative change in determined Adjusted R-squared for the identified models.

Model	Adjusted R^2	Absolute Change from Base Model	Relative Change from Base Model
Multiple Linear Regression (Base Model; Table 1)	20.03%
MARS – Additive (Table 2)	22.47%	2.44%	12.18%
MARS – Interaction (Table 3)	29.77%	9.74%	48.63%

Figure 3. a) Alcohol-based hand rub chart showing the statistical significance (t-values) of the estimated coefficients related to the threshold values from the OAT sensitivity analysis. The solid black vertical lines mark an identified threshold. b) Third-generation cephalosporin chart showing the statistical significance (t-values) of the estimated coefficients related to the threshold values from the OAT sensitivity analysis. The solid black vertical lines mark an identified threshold.

Figure 4. Contribution chart for the effect of alcohol-based hand rub (ABHR) on the incidence rate of ESBL-producing K. pneumoniae holding third-generation cephalosporin use at constant levels..