Nephrology Comanagement and the Quality of Antibiotic Prescribing in Primary Care for Patients With Chronic Kidney Disease: A Retrospective Cross-sectional Study

A Retrospective Cross-Sectional Study

Justin X.G. Zhu; Danielle M. Nash; Eric McArthur; Alexandra Farag; Amit X. Garg; Arsh K. Jain


Nephrol Dial Transplant. 2019;34(4):642-649. 

In This Article


Study Patients and Baseline Characteristics

After exclusions there were 13 875 eligible patients with a study antibiotic prescription from a primary care physician. Of these, 5961 (43%) patients were comanaged by a nephrologist. In the comanaged group, the most recent nephrologist visit was a median of 72 (25th–75th percentile 34–135) days prior to the antibiotic prescription date.

After matching, we retained 3937 unique patients in each group for a total of 7874 patients. Patient selection is presented in Figure 1. The two groups were well balanced across baseline characteristics after matching (Table 2). Patients had a median age of 81 (25th–75th percentile 76–86) years and 63% were female. Patients had a median eGFR of 25 (25th–75th percentile 21–28) mL/min/1.73 m2 and 94% of the patients had Stage 4 CKD. The number of patients with two or more serum creatinine measurements in the year prior to the index date was 6049 of 7874 patients (76.8%). Approximately 11% of patients resided in a rural location and 9% of patients were in a long-term care facility.

Figure 1.

Participant flow diagram.

Antibiotic Prescriptions

Of the 11 study antibiotics, cephalexin, ciprofloxacin and clarithromycin accounted for 55% of all prescriptions (21, 18 and 16%, respectively). There were no differences in terms of frequency of antibiotic prescriptions between noncomanaged and comanaged groups except for nitrofurantoin (Table 3). Nitrofurantoin (which is contraindicated in advanced CKD) was more frequently prescribed in patients in the noncomanaged group.

In total, the overall percentage of appropriately dosed prescriptions for the study cohort was 32% (2526/7874). Figure 2 depicts the number of inappropriately dosed prescriptions by the type of antibiotic. Cefixime and ofloxacin were grouped together due to small numbers and accounted for 1% of total prescriptions.

Figure 2.

Number and percentage of appropriately and inappropriately dosed prescriptions by antibiotic type. Percentages above each bar denote the percentage of appropriately dosed prescriptions for that antibiotic.

Association of Comanagement and Appropriately Dosed Prescriptions

In the absence of nephrology comanagement, 1184/3937 (30%) patients had appropriate doses of an antibiotic. Nephrology comanagement was associated with an increase in the chance that an appropriate dose of an antibiotic was prescribed by a primary care physician, although the effect was modest {1342/3937 [34%]; OR 1.20 [95% confidence interval (CI) 1.09–1.32]; P < 0.001} (Table 4). This corresponded to an absolute difference of 4.0% (95% CI 2.0–6.1) between the groups.

Additional Analyses: Effects of Mandatory eGFR Reporting or the Degree of Renal Dysfunction

Mandatory eGFR reporting on laboratory reports (instituted in January 2006 in Ontario) did not significantly modify the association between comanagement and appropriately dosed prescriptions (interaction P = 0.08) (Table 5). The prevalence of total inappropriate antibiotic prescriptions in both groups combined was 66% pre-2006 and continued to be high post-2006 (69%).

A second subgroup analysis was performed to examine if the association of comanagement with appropriately dosed prescriptions was modified by stage of CKD: CKD Stage 4 (eGFR 15–<30 mL/min/1.73 m2) versus CKD Stage 5 (eGFR <15 mL/min/1.73 m2) (Table 5). There was no evidence of interaction by CKD stage with a conditional OR of 1.19 (95% CI 1.08–1.31) in Stage 4 CKD and 1.37 (95% CI 0.96–1.97) in Stage 5 CKD (interaction P = 0.5) (Table 5).

Additional Analyses: Accounting for Prescriptions by the Same Physician

A total of 7874 patients were under the care of 3461 unique primary care physicians. Of 3461 physicians, 1315 cared for patients belonging to both comanaged and noncomanaged groups, yielding a total of 2494 in the comanaged group and 2282 in the noncomanaged group. In additional analyses, accounting for the correlation of patients followed by the same primary care physician, the results were the same to two decimal points. The 3937 patients were comanaged by 220 unique nephrologists.