Are Antibiotic Drugs Well Prescribed in Case of Renal Insufficiency? A Retrospective Study

Abstract

Renal insufficiency (RI) is a major complication in hospitalized patients. We aim to determinate if the severity of RI is considered for antibiotic prescriptions. A 10-month retrospective study including all in-patients of an orthopedic surgery department, based on the analysis of antibiotic prescriptions of patients with RI, was set up as follows: identification of patients with RI estimated with Cockcroft formula, classification by severity stage, and analysis of antibiotic prescriptions to be adapted to RI. About 10% of patients had RI. Among them, 54 (32%) received antibiotics (on average, 1.75 drugs per patient). Sixteen (17%) of antibiotic prescriptions required either dose adaptation or therapeutic drug monitoring. In all, only four prescriptions were adapted to renal function. In other cases, antibiotics were prescribed according to protocols for patients with normal renal function. Moreover, therapeutic drug monitoring was only performed for half of required cases and then showed values > ULN three times out of four. Creatinine clearance (CrCl) has been calculated for half of patients with RI. In practice, dosage adjustment of antibiotics is done only for patients with severe RI. Within the framework of the introduction of an electronic prescribing technology and medication order pharmaceutical review procedures, CrCl is now systematically calculated and then taken into account by both prescribers and clinical pharmacists.

Keywords:

• antibiotics
• clinical pharmacy
• creatinine clearance
• renal insufficiency

INTRODUCTION

Renal insufficiency (RI) is a common type of organ failure that often remains underdiagnosed, which can result in increased morbidity, mortality, and health care costs. Clinical pharmacists are involved in promoting appropriate antibiotics use for RI patients. It has been reported that about 20% of hospital pharmacists' interventions concern drugs dosage adjustment in case of RI,[1] and the problem will grow with the general aging of the population. Drug dose adjustment is of importance to prevent renal toxicity, especially considering drugs with narrow therapeutic margin and compounds known for their renal toxicity, such as antibiotics. Therefore, the purpose of this study was to determine if renal parameter (i.e., creatinine clearance, CrCl) was measured routinely before antibiotics prescription, and then if dosage adjustment was performed when required, according to pharmacokinetic and pharmacodynamic recommendations, which might make the prescription questionable in those high-risk patients.

MATERIALS AND METHODS

We performed a 10-month retrospective study including all in-patients of one unit of the orthopedic surgery department of a large university hospital. Indeed, this unit is included in a clinical pharmaceutical analysis and validation process experiment, providing clinical pharmacists several variables to calculate CrCl according to Cockcroft-Gault formula:[2]

A correction factor of 0.85 was applied for females. We collected all treatment lines that each patient received during his or her hospitalization, and, when antibiotics prescriptions were available, we collected the name of the drug, the dosage, the route of administration, and the duration of treatment. We identified patients with RI according to CrCl estimated with the Cockcroft-Gault formula, and we classified them by severity stages according to the National Kidney Foundation Practice Guidelines, based on a systematic review of the literature.[3] We identified both patients with RI for whom antibiotics were prescribed and investigated actual dose adaptation according to the stage of RI.

RESULTS

In all, 1,703 patients were included into the study; of these, 169 (about 10%) suffered from RI. Among those RI patients, 54 received antibiotics. The ratio male/female was 17/37, the mean age was 73.9 years old (range 44–98), and the mean weight 65.5 kg. Patient's profile according to RI severity showed 46% with slight RI (CrCl between 60–89 mL/min), 41% with moderate RI (CrCl between 30–59 mL/min), 11% with severe RI (CrCl between 15–29 mL/min), and 2% with major RI (CrCl <15 mL/min, or dialysis). The main antibiotic drugs prescribed were the penicillin-beta-lactamase inhibitor association amoxicillin + clavulanic acid (23%), rifampicin (14%), and several fluoroquinolones (i.e., ofloxacin, 16%; norfloxacin, 14%; and cloxacillin, 4%).

The mean number of antibiotics administrated to RI patient was 1.75 (range: 1–4) calculated from 96 antibiotic prescriptions. Sixteen of them required either dosage adjustment or therapeutic drug monitoring; in fact, 12 antibiotic prescriptions from 16 were not adapted to renal function. More precisely, dosage adjustment has not been performed for cotrimoxazole (one prescription), teicoplanin (one prescription), vancomycin (one prescription), and several fluoroquinolones (ofloxacin and ciprofloxacin, four prescriptions). Moreover, therapeutic drug monitoring has not been performed for rifampicin (two prescriptions), and teicoplanin (one prescription). When it was available, therapeutic drug monitoring showed blood concentration values were greater than the upper limit of normal for three patients with rifampicin (two prescriptions) and vancomycin (one prescription). None of the concerned patients developed clinical significant adverse effects related to their antibiotics prescription.

DISCUSSION

Chronic RI is a worldwide public health problem with increasing incidence and prevalence and high cost, especially in hospital. Antibiotics, with ACE inhibitors and nonsteroidal anti-inflammatory drugs (NSAID), constitute one of the drug classes that is most commonly involved in a short-term decline of renal function.[4] The pharmaceutical care of RI patients requiring antibiotic therapy requires a good correlation of the patient's renal function status with the selected antibiotic. In our study, we found that dose adjustment is made in a minority of cases. This result may point out the role the clinical pharmacist has to play in the fight against inappropriate antibiotics prescription, as has been previously described,[5] and illustrate the wide range of actions the pharmacist has at his disposal in hospitals to diminish drug-related problems (DRPs), of which many may be preventable due to their predictability.[6]

Even if appropriate dosages are determined according to the varying degrees of RI, only severe RI is taken into account at the time of antibiotic prescriptions, which may be due to greater caution being used in this high-risk population. However, a low number or patients (i.e., seven) with severity stage ≥ 4 were admitted to the study, which may partly explain this fact. In our study, every patient had clinical complication due to an inappropriate use of antibiotics, but the use of renal high-risk drugs in hospitalized patients with RI may be harmful and have deleterious effects: for instance, central nervous system neurotoxicity due to penicillin and beta-lactam antibiotics is a documented DRP.[7]

The discrepancy between required dose adjustments and prescribers' practice may be in part the result of the method by which renal impairment was assessed. The point then becomes obtaining a more efficient estimation of the RI severity: CrCl has been calculated for only half of patients. In practice, dosage adjustment is done according to serum creatinine, which is an indirect and not powerful index of renal function:[8] depending on the patient's physical variables, a minor elevation of serum creatinine, indeed remaining in the normal range, may be consistent with a substantial decline of renal function. To estimate the severity degree of renal disease, CrCl, one of the best overall indices of the kidney function, is a more efficient marker and constitutes a validated tool to measure RI when performed according to patient's physical variables (weight, sex, and age). Nevertheless, in monitoring a large number of patients, serum creatinine use appears to be unavoidable despite the disadvantage of his marker. In this context, the calculation of the CrCl value with computerized prescribing software in our teaching hospital would improve care of chronic RI patients. Besides, even if none of the antibiotic drugs listed in this paper requires statutory therapeutic drug monitoring, it is noteworthy that antibiotics should be monitored by analysis of serum drug levels when available from the pharmacological laboratory of the hospital, as this is the most effective confirmation of safe therapy.

We reviewed the pharmacokinetics of all antibiotic drugs prescribed in those patients, according to available data from international literature.[9],[10] It is noteworthy that recommendations for the choice of an antibiotic drug in patients with RI should be based on many factors, such as the bacterial species considered, the localization of the infectious seat, the possibility to adjust dosage or to increase in the dosing interval according to the severity of RI, and the renal tolerance of the drug. Antibiotic drugs were thus sorted according to their pharmacokinetic behavior in RI and their renal tolerance profile.

The pharmacokinetics of cloxacillin,[11] oxacillin,[12] pefloxacin,[13] ceftriaxone,[14] and fusidic acid[15] should not be influenced by RI. In renal impairment (including patients with end-stage renal failure), these drugs are reported to be safe and effective. Thus, dosing adjustment is unlikely to be necessary in patients with pre-existing renal impairment. With norfloxacin[13] and roxithromycin,[16] neither reduction of the doses nor therapeutic drug monitoring is necessary in patients with mild to severe renal impairment except for patients with terminal RI (CrCl < 15 mL/min). Then, half of standard dosage is recommended. Dosages of amoxicillin,[17] ciprofloxacin, ofloxacin,[13] clindamycin,[18] and cotrimoxazole (trimethoprim + sulfamethoxazole)[19] have to be adjusted according to the severity of RI: a progressive decrease of the dosage is recommended with these agents. With rifampicin,[20],[21] no dose adjustment is required in patients with different degrees of RI; however serum rifampicin concentrations should be monitored closely after initiating treatment, especially in this high-risk population. Standard peak serum rifampicin concentrations range from 8–12 mg/L in the case of tubercular infection and from 18–30 mg/L in case of Gram+ infection. Teicoplanin[22] and vancomycin[23] require both dosage adjustment and therapeutic drug monitoring in case of RI. Indeed, with these two drugs, the efficient peak would be attained earlier in patients with RI when compared to patients with normal renal function. It is thus mandatory to closely monitor those patients. It is recommended that unitary doses of teicoplanin should be reduced in patients with impaired renal function (half of standard dosage when CrCl < 60 mL/min). Standard peak serum teicoplanin concentrations range from 15–20 mg/L. To improve vancomycin safety, it has to be prescribed using an extended dosing interval with normal unitary doses. Standard peak serum vancomycin concentrations range from 20–40 mg/L and less than 20 mg/L in continuous perfusion. All of these dosing recommendations for adult patients with RI are summarized in Table 1.

Table 1 Dosing recommendations for adult patients with renal insufficiency[10]

Drug	Creatinine clearance (mL/min)
	90–60	59–30	29–15	< 15 + hemodialysis
Amoxicillin	1000 mg/12 h	1000 mg/12 h	500 mg/12 h	500 mg/24 h
Amoxicillin + clavulanic acid	1000 mg/12 h	1000 mg/12 h	500 mg/12 h	500 mg/24 h
Ceftriaxone	1000–2000 mg/d	1000–2000 mg/d	1000–2000 mg/d	1000–2000 mg/d
Ciprofloxacin	250–1500 mg/d	500 mg/d	500 mg/d	500 mg/d
Clindamycin	600–2400 mg/d	600–2400 mg/d	600–2400 mg/d	600–1200 mg/d
Cloxacillin	Per os: 25–50 mg/kg/d in two doses, IV: 50–100 mg/kg/d	Per os: 25–50 mg/kg/d in two doses,IV: 50–100 mg/kg/d	Per os: 25–50 mg/kg/d in two doses,IV: 50–100 mg/kg/d	Per os: 25–50 mg/kg/d in two doses, IV:50–100 mg/kg/d
Cotrimoxazole	800 mg/12 h	800 mg/12 h	800 mg/24 h	800 mg/48 h
Fusidic acid	1000–1500 mg/d	1000–1500 mg/d	1000–1500 mg/d	1000–1500 mg/d
Norfloxacin	800 mg/d	800 mg/d	800 mg/d	400 mg/d
Ofloxacin	400–600 mg/d	200 mg/d	200 mg/48 h	200 mg/48 h
Oxacillin	Per os: 50–100 mg/kg/d in two doses, IV: 35–50 mg/kg/d	Per os: 50–100 mg/kg/d in two doses,IV: 35–50 mg/kg/d	Per os: 50–100 mg/kg/d in two doses,IV: 35–50 mg/kg/d	Per os: 50–100 mg/kg/d in two doses,IV: 35–50 mg/kg/d
Pefloxacin	800 mg/d	800 mg/d	800 mg/d	800 mg/d
Pristinamycin	2000–3000 mg/d	2000–3000 mg/d	2000–3000 mg/d	2000–3000 mg/d
Rifampicin	10–20 mg/kg/d	10–20 mg/kg/d	10–20 mg/kg/d	10–20 mg/kg/d
Roxithromycin	5–8 mg/kg/d in two doses	5–8 mg/kg/d in two doses	5–8 mg/kg/d in two doses	2.5–4 mg/kg/d in two doses
Teicoplanin	200–400 mg/d	100–200 mg/d	100–200 mg/d	100–200 mg/d
Vancomycin	500–1000 mg/d	500–1000 mg/d	500–1000 mg/d	500–1000 mg/d

In this context, the clinical pharmacist provides an important contribution in the multidisciplinary hospital team, not only by focusing upon the economic impact generated by drug prescription procedures, but also by promoting patient safety and preventing or reducing risk arising from DRPs, and enhancing the efficiency of drug treatment. According to a recent study on dose-related problems and the need for laboratory tests, both of the DRPs identified in our work are important, requiring clinical pharmacist-targeted interventions.[24] A validation procedure of hospital prescriptions by clinical pharmacist may contribute to diminishing DRPs, help modify common attitudes toward prescribing antibiotics, and then improve drug-related morbidity and mortality.

This study may be criticized due to the following methodological limitations. First, we performed a retrospective analysis, which should be validated by a prospective study. Secondly, we did not work on aminoglycosides, the most nephrotoxic antibiotic drug class, because any patients with identified RI received these drugs, pointing to its safe and restrained prescription in those high-risk patients. Instead, we focused on the most commonly prescribed antibiotics in the case of orthopedic surgical procedure according to local protocols, especially total hip replacement, open reduction, and internal fixation of the femur to demonstrate for clinicians efficient guidelines for their daily practice. As an aside, we did not take into account the fact that anesthesia and surgery may influence physiological serum creatinine level and might indicate an overestimation of RI,[25] as well as the coadministration of other nephrotoxic drugs, especially NSAID, which may transiently affect globular and tubular functions.

At the end of that work, and within the framework of a computerized prescription procedure and pharmaceutical validation, CrCl is now systematically calculated for detecting RI and classifying its severity. Prescribers have also been made sensitive to take it into account as an effective tool to manage their prescriptions in order to improve patient safety. The outcome of this advancement on antibiotics prescriptions will be evaluated within a global before/after study about medicine network's computerization in our teaching hospital.

Notes

9. ANAES Department of Professional Guidelines. Chronic renal failure diagnosis in adult. ANAES, St. Denis La Plaine, France; 2002.