Why Bother? Lab Monitoring in Beta-Lactam Outpatient Parenteral Antimicrobial Therapy

LETTER

The 2018 Infectious Diseases Society of America guidelines on outpatient parenteral antimicrobial therapy (OPAT) recommend a weekly complete blood cell count and complete metabolic panel for most antimicrobial therapies, and this weekly lab monitoring is standard practice among OPAT providers (1, 2). However, few studies address the effectiveness of laboratory monitoring in OPAT or comment directly on the utility of laboratory values for patient care decisions (3–7). Therefore, we investigated the real-world clinical value of weekly laboratory monitoring for patients in our OPAT program, i.e., how often the laboratory monitoring led to therapy modification. We focused on patients receiving beta-lactam OPAT, since beta-lactams are some of the best tolerated OPAT medications, and therefore the necessity of weekly lab monitoring is perhaps debatable.

We reviewed the charts of all adult patients discharged from University of Utah Health between 1 January 2018 and 31 July 2019 who received monotherapy beta-lactam OPAT from a home health service; had planned follow-up care with our Infectious Disease division; were expected to be on OPAT for ≥7 days; and had at least one set of OPAT labs collected. For each patient, we determined if (i) they had any abnormal laboratory values during OPAT and (ii) whether that abnormal value led to a change in therapy. Laboratory values were defined as abnormal and concerning for a potential adverse effect based upon prespecified criteria: creatinine increase by ≥0.3 mg/dL or ≥50% (8, 9); aspartate/alanine aminotransferase or bilirubin at ≥110% upper limit of normal (10); white blood cell count of ≤4,000/mm³ (11); absolute neutrophil count of ≤1,500/mm³; platelets of <100/mm³ (10); and eosinophils of ≥500/mm³ (12). Anemia and leukocytosis were not considered, since the majority of patients were discharged with a baseline anemia and/or leukocytosis. We reviewed lab values from the day of discharge until 6 weeks after discharge or until the end of OPAT, whichever was sooner. Alterations in antimicrobial choice, dose, or duration were considered therapy modifications. The Institutional Review Board deemed our project exempt from formal review.

We reviewed 264 patient records. Most patients were white (88%) and male (66%), with an average age of 53 years (standard deviation, 16 years). Cephalosporins were the most prescribed beta-lactams for OPAT (67%). Abnormal laboratory values were identified in 137 patients (52%) (Fig. 1). Ninety-three patients (35%) with abnormal laboratory values were maintained on their OPAT regimen without a modification. Another 29 patients (11%) with abnormal labs had their therapy changed, but it was not due to the abnormal lab value. For only 15 patients (5.7%) did the abnormal OPAT lab lead to a therapy modification (Table 1). Another two patients had therapy modified due to lab values that did not meet our criteria for abnormal (e.g., dose adjusted for improving renal function). In total, 17 patients (6.4%) had their OPAT therapy modified due to their weekly OPAT lab results.

FIG 1.

Waffle chart of patient population. One circle, 1%.

TABLE 1.

Patients with therapy modifications due to abnormal laboratory values

Patient	OPAT antibiotic	Change	Reason	Timinga
1	Cefazolin	Changed to ceftriaxone	Increased creatinine	12
2	Cefazolin	Changed to ceftriaxone	Increased creatinine	5
3	Cefazolin	Changed to levofloxacin	Increased creatinine	18
4	Cefazolin	Changed to levofloxacin	Neutropenia	22
5	Cefazolin	Changed to levofloxacin	Thrombocytopenia	33
6	Cefazolin	Changed to vancomycin	Hepatotoxicity	21
7	Cefazolin	Changed to levofloxacin	Cytopenia	20
8	Cefepime	Dose reduced	Increased creatinine	22
9	Cefepime	Changed to levofloxacin	Cytopenia	12
10	Cefepime	Changed to levofloxacin	Leukopenia	25
11	Cefepime	Dose reduced	Increased creatinine	22
12	Ceftriaxone	Changed to ampicillin-sulbactam	Hepatotoxicity	26
13	Penicillin	Changed to ceftriaxone	Increased creatinine	12
14	Piperacillin-tazobactam	Changed to ertapenem	Increased creatinine	19
15	Nafcillin	Antibiotics stopped	Eosinophilia	28

^aDay of OPAT that the change occurred.

Not surprisingly, patients who had abnormal laboratory values during OPAT were older (55 years versus 51 years, P = 0.03) and sicker (median Charlson comorbidity index, 3 versus 2, P = 0.04) than those who did not have abnormal laboratory values. Patients with abnormal lab values were more likely to have diabetes or kidney disease or be receiving chemotherapy. However, we were unable to identify categories of patients more likely to have their therapy modified due to an abnormal result and who would thus benefit most from laboratory monitoring. Further, therapy modifications occurred for most adverse drug reactions of interest (hepatotoxicity, nephrotoxicity, and cytopenia) and throughout the course of therapy (range, <1 week to >4 weeks). Therefore, we are unable to recommend narrowing the scope of laboratory monitoring for beta-lactam OPAT.

While OPAT modifications due to abnormal laboratory values were uncommon, abnormal laboratory values were exceedingly common: half of patients had at least one result concerning for a potential adverse event. Since the clinician must consider whether each of these abnormal results requires action, OPAT lab monitoring takes significant time and mental energy. However, abnormal values seldom lead to action. Our impression from chart review is that most abnormal values during OPAT are not far enough outside the normal range to be worrisome. In addition, patients often have other explanations for their abnormal results (other medications or chronic disease).

Based on our findings, we conclude that weekly lab monitoring during OPAT is necessary to capture rare but serious adverse effects. However, we see an opportunity to improve the efficiency of weekly laboratory monitoring by reducing provider time spent evaluating abnormal results that ultimately have no clinical significance. At present, all abnormal results are flagged similarly by electronic medical record systems. Future research could identify which patterns of results are considered clinically significant by OPAT clinicians and create algorithms to flag those results differently. For example, a platelet count that has slowly drifted down from 230 to 90 while on OPAT is more concerning than one that has dropped from 100 to 90. Using more sophisticated electronic algorithms could therefore improve the efficiency of OPAT lab monitoring.