The impact of procalcitonin-guided antibiotic treatment on treatment duration and hospital costs in children—balancing benefits and burdens

A growing body of high-quality evidence supports shorter courses of antibiotics for serious bacterial infections. Despite antimicrobial stewardship efforts, antimicrobials remain one of the most commonly prescribed medications in children, accounting for approximately 30–45% of prescriptions in hospitalised children in high-income countries (1,2). Although detection of a pathogen by either culture and/or polymerase chain reaction remains the gold standard for diagnosis of bacterial infections, only 9% of sepsis episodes in children are culture positive (3). As a result, biomarkers such as procalcitonin (PCT) have been proposed to provide an alternative means to differentiate viral from bacterial infections.

PCT is a prohormone whose production is upregulated in response to lipopolysaccharide (LPS) and other pro-inflammatory mediators present in bacterial infection. PCT is detectable in serum within 2–4 hours of LPS exposure with peak levels achieved within 12 hours (4). In contrast, serum PCT is not increased in viral infections, as high levels of interferon-γ suppress the upregulation of CALC1 gene expression responsible for PCT production (4). With appropriate antimicrobial treatment, serum levels of PCT rapidly decrease due to its relatively short half-life of approximately 24 hours, therefore offering a possible means of monitoring response to treatment (4). In 2015, the National Institute for Health and Care Excellence (NICE) guideline reviewed 18 studies, of which only two were randomised controlled trials (RCTs) in children (with low-certainty evidence) and found there was inadequate evidence to support the use of PCT in routine clinical care (5). Despite this, guidance published by the American Academy of Pediatrics in 2021 recommends the use of a PCT level in the initial evaluation of all febrile infants aged 8–60 days. However, there are currently no recommendations for the use of PCT to guide treatment duration (6).

In the recent multinational UK BATCH RCT published in The Lancet Child & Adolescent Health, Waldron and colleagues aimed to address this research gap in PCT-guided decision making in children (7). The trial enrolled 1,949 children aged 72 hours to 18 years old across 15 hospitals who required short-term intravenous (IV) antibiotics for suspected or confirmed bacterial infections, including lower respiratory tract infections (LRTI) (22%), gastrointestinal or abdominal infections (15%), soft tissue infection (14%) and sepsis (13%). Participants were randomised to receive either usual clinical management without PCT measurement (972, 49.9%) or PCT-guided antibiotic treatment (977, 50.1%) with PCT repeated at 1 to 3-day intervals. In the latter group, clinicians were advised to cease antibiotics if the initial PCT within 48 hours was <0.25 ng/mL. In cases of suspected infection where the initial PCT was ≥0.25 ng/mL and those with confirmed infection, clinicians were encouraged to cease IV antibiotics after 48 hours if the PCT was ≤0.5 ng/mL or if PCT decreased by ≥80% of the peak and was ≤1 ng/mL. The trial found that PCT-guided treatment did not reduce the duration of IV antibiotic treatment compared to usual clinical management [PCT median 99.7 hours, interquartile (IQR) 61.2–153.8 hours vs. usual management median 96.0 hours, IQR 59.5–155.5 hours; hazard ratio (HR) 0.95, 95% confidence interval (CI): 0.87–1.05] nor time to hospital discharge or total antibiotic duration. The overall lack of a difference in clinical outcomes between the two groups may be partly explained by only one-quarter (24%) of participants in the PCT-guided group having PCT results considered at all reviews. Notably, in this subgroup, a shorter duration of antibiotics was observed [PCT median 77.3 hours (IQR 50.0–132.0 hours) vs. usual management median 99.7 hours (IQR 61.2–153.8 hours), P value not stated]. Also, critically, strict adherence to the PCT-guided treatment algorithm for the duration of antibiotic treatment occurred in only 16% of patients. This trial highlights the practical challenges of PCT-guided treatment but the results suggest that in centres where PCT levels are accessible and clinicians comply with incorporating PCT levels consistently in their decision making, it might play a role in reducing the duration of antibiotic treatment.

PCT-guided antibiotic treatment in children has been evaluated in three previous RCTs, involving a total of 913 children (8-10). In two of these trials, PCT monitoring was found to reduce antibiotic treatment duration with no associated reduction in duration of hospital stay (8,9). The largest RCT enrolled 337 children aged 1 month to 18 years with LRTI and found a significant reduction in mean antibiotic duration (4.5 vs. 6.3 days, P=0.04) but no difference in antibiotic prescription rates or duration of hospital stay in the PCT-guided group using a PCT-threshold of 0.25 ng/mL for cessation of antibiotics, compared with standard care (8). This contrasts the results of an RCT involving 270 paediatric intensive care unit (PICU) patients that used a higher PCT-threshold of 0.5 ng/mL to de-escalate antibiotic treatment and found no difference in median cumulative antibiotic days of treatment or total antibiotic treatment duration compared to standard care (9). The results of this trial are consistent with those of the BATCH trial, despite higher clinician adherence to the PCT algorithm in the PICU trial (BATCH 16% vs. PICU 70%). These three studies were included in a recent meta-analysis by Li et al. of four paediatric RCTs involving 1,313 children which found a significant reduction in antibiotic treatment duration (−2.2 days) and antibiotic-related adverse events (relative risk 0.25) but no difference in duration of hospital stay, antibiotic prescription rate, hospital readmission or mortality (11). However, there was a high level of heterogeneity (I²=79–93%) in three of the five outcomes evaluated in the meta-analysis. A possible reason for this is the use of different PCT thresholds (0.25 and 0.5 ng/mL) for initiating or continuing antibiotic treatment, the latter consistent with the BATCH trial. Also, the included studies enrolled children with either LRTI or sepsis, contrasting the heterogeneous infections included in the BATCH trial (only 36% with LRTI or sepsis). Previous observational studies have shown that the specificity of PCT for diagnosing bacterial infections varies with the site of infection (90% for bone and joint infections, 88% for LRTI, 76% for sepsis) with limited data for ear, nose and throat infections (12-14). This heterogeneity may have contributed to the difference in findings between this systematic review and the BATCH trial. Studies including immunocompromised children are limited, although one small single-centre RCT enrolled 46 low-risk febrile neutropenic children receiving chemotherapy and reported a significant reduction in median antibiotic treatment duration (3 vs. 7 days, P<0.001) but no difference in antibiotic treatment failure using PCT-guided treatment duration compared with a standard 7-day course (15). Further studies in this patient group are required to validate these results.

PCT-guided antibiotic treatment has been extensively investigated in adults. A recent systematic review and meta-analysis by Papp et al. included 9,048 adults with suspected or confirmed bacterial infection across 26 RCTs and found PCT-guided treatment resulted in a significant reduction in duration of antibiotic treatment (mean difference: −1.79 days, 95% CI: −2.65 to −0.92) and 28-day mortality [odds ratio (OR) 0.84, 95% CI: 0.74 to 0.95] but no significant change in duration of hospital or intensive care unit (ICU) stay (16). These findings were largely informed by the SAPS trial which randomised 1,546 adults in the Netherlands and found a significant decrease in both 28-day and 1-year mortality (35% vs. 41%, absolute difference 6.1% [95% CI: 1.2% to 10.9%], P=0.02) in the PCT group compared with standard care (17). In the recent ADAPT-Sepsis multicentre RCT, which enrolled 2,760 adults with suspected sepsis from 41 UK ICUs, PCT-guided treatment significantly reduced antibiotic duration (9.8 vs. 10.7 days, P=0.01) compared to standard care (18). Interestingly, the authors found a non-inferior increase in 28-day all-cause mortality in the PCT group [20.9% PCT vs. 19.4% standard care; absolute difference 1.6% (95% CI: −2.2% to 5.3%)] but no difference in 90-day all-cause mortality. This lack of mortality benefit is consistent with the paediatric data to date, perhaps due to the lower mortality from sepsis in children compared to adults.

Access to and cost of PCT assays in comparison to C-reactive protein (CRP) are often mentioned as barriers to the implementation of PCT-guided treatment. PCT is a costly assay due to the requirement for specialised equipment and reagents. Costings from the Netherlands, UK and Colombia reveal that PCT assays are approximately ten times more expensive than CRP assays (PCT £14–50 vs. CRP £1–7) (7,19-22). In the recent UK BATCH trial in children, PCT-guided treatment was therefore found not to be a cost-effective strategy. The adult SAPS trial, however, found no significant difference in mean hospital costs between the PCT group compared to standard care [€46,081 vs. €46,146 (95% CI: €−6,314 to €6,107)] (19). In fact, a non-significant increase in healthcare costs over a one-year period was observed [PCT mean €73,665 vs. standard care mean €70,961 (95% CI: €−4,495 to €10,005)], likely due to the higher healthcare-related costs in sepsis survivors after ICU discharge compared to baseline in the PCT group [monthly cost difference €2,281 (95% CI: €−6,314 to €6,107), P<0.001] (20,23). These findings show that the cost-effectiveness of PCT-guided treatment is closely linked to its impact on mortality, though this has not been demonstrated in children.

Overall, current evidence from studies in children does not support the incorporation of PCT testing into routine clinical practice to guide antibiotic treatment duration. Although findings from two RCTs suggest that it may reduce antibiotic treatment duration, there is insufficient evidence of clinical benefit in terms of reduced duration of hospital admission, admission to intensive care or mortality. Moreover, the increased costs associated with PCT testing are not offset by cost savings through reduced duration of hospital stay or antibiotic use.

Supplementary

The article’s supplementary files as

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.