Prevalence of Culture-Negative Fever in Infants With Down Syndrome Undergoing Cardiac Surgery

Introduction

Nearly half of infants born with Down syndrome (DS) have congenital heart disease—frequently of the type that will require cardiac surgery. ¹ Among non-single ventricle lesions, patients with DS have similar mortality rates compared to their nonsyndromic counterparts. ² Yet, they do demonstrate greater morbidity and complication rates including sepsis and other infections. ^2,3 This is thought, in part, to be due to a baseline pro-inflammatory state and abnormal immune response, as well as improper activation of circulating lymphocytes. ^4,5

Even without overt sepsis or demonstrated infection, infants with DS seem to have higher rates of culture-negative fever after cardiac surgery. While rates of documented infection following surgery, especially bacteremia and respiratory infection, have been well-documented, rates of fever alone without clinical or laboratory evidence of infection remain poorly described. The objective of this study was to compare the prevalence of culture-negative postoperative fever in infants with and without DS and explore how the presence of fever affected the operative recovery among patients with DS.

Methods

A retrospective case–control study was performed to evaluate the outcome of interest. Between 2010 and 2016, all patients less than one year of age with DS who underwent surgery at our institution for an atrioventricular septal defect or ventricular septal defect were included. Control group encompassed children less than one year of age without chromosomal anomalies who were age and surgery matched to the patients with DS. Control subjects were added in reverse time order, starting with the most recent, until a 1:1 match was created. The study period was selected secondary to reliability of temperature measurements documented in the electronic medical record at our institution.

Patients were excluded if they had a positive blood, urine, or wound culture during the admission that could be an etiology for the fever. Patients were also excluded if they required extracorporeal membrane oxygenation secondary to the significant inflammatory response that occurs with use of this support. Controls were excluded, in addition to the above, if any chromosomal anomalies were determined at the time of review. Patients with conotruncal defects including tetralogy of Fallot, truncus arteriosus, and interrupted aortic arch were not included due to the high association with 22q.11 deletions. If, during analysis, a control patient was eliminated due to exclusion criteria, the next most recent patient was chosen to minimize selection bias.

Temperatures were recorded for the first 72 hours after surgery. A fever was defined as a rectal temperature greater than 38.5°C, which was the standard in the cardiac intensive care unit (ICU) during the study period. In an attempt to accurately define fever over this time period and characterize both maximum temperature and duration, an area under the curve (AUC) method was utilized (Figure 1). ⁶ The AUC equation used a temperature of 38.5°C as the minimum fever line.

OPEN IN VIEWER

Figure 1.

Example graph of a fever curve during the first 72 hours after cardiac surgery. The area under the curve (AUC) was calculated for each patient and compared between groups.

Data collected included demographics, temperatures for the first 72 hours after surgery, length of stay (LOS), hospital outcomes, and costs. Financial data, including hospital cost and charge, were gathered from the internal hospital financial database and reported as 2016 dollars adjusted appropriately for inflation. Cost is defined from the standpoint of the medical institution and is created as a fixed ratio of charge data based on the environment the service has been provided. Continuous, normally distributed data are reported as mean ± standard deviation, unless noted, while categorical variables are reported as frequencies (%). Comparisons were made using t-test or χ², as appropriate. A P value less than .05 was considered significant. The institutional review board at the University of Virginia approved this study.

Results

There were 45 patients identified who had DS and met the inclusion criteria with 45 controls identified. Demographics and outcomes for both groups are shown in Table 1. There were no differences in baseline demographics including age, weight, and gender distributions. Patients with DS did have longer time on bypass, a longer postoperative recovery time including ventilator time, longer ICU LOS, and longer total LOS compared to controls. There was also a higher financial burden on the DS group as they accumulated over 50% higher hospital costs and charges compared to the control population. Patients with DS who developed fever demonstrated no difference in their white blood cell (WBC) count at the time of fever compared to control patients who also developed fever.

OPEN IN VIEWER

Table 1.

Comparisons of Demographics, Surgical Variables, Outcomes, and Costs Between Patients With DS and Those Without.^a

	Down Syndrome (n = 45)	Control (n = 45)	P Value
Age, days	128.9 ± 59.9	123.0 ± 80.0	.72
Age range, days	26-320	28-295
Weight at surgery, kg	4.9 ± 1.1	4.9 ± 1.3	.98
Female, n (%)	19 (42%)	21 (47%)	.81
CPB time, minutes	124.3 ± 67.0	84.4 ± 48.5	<.01
Ventilator time, hours	64.1 ± 60.4	23.6 ± 31.9	<.01
ICU LOS, days	6.3 ± 4.1	4.1 ± 2.9	.01
Total LOS, days	10.4 ± 7.1	6.6 ± 4.2	.01
Total hospital cost, US$	45,616 ± 23,241	30,025 ± 13,394	<.01
Total hospital charge, US$	146,691 ± 70,205	88,009 ± 36,289	<.01
Doses of antipyretics (first 72 hours)	6.9 ± 3.5	9.0 ± 3.4	.02
WBC at time of first fever, if present, k/µL	12.3 ± 2.3	14.1 ± 3.9	.24

Abbreviations: CPB, cardiopulmonary bypass; DS, Down syndrome; ICU, intensive care unit; LOS, length of stay; WBC, white blood cell count.

^aData are presented as mean ± standard deviation or n (%).

The overall prevalence of fever was higher in the DS group compared to the control group (58% vs 24%, P = .003), but there was no difference in the mean AUC greater than 38.5°C between the two groups (4.9 vs 4.1, P = .57). The importance being that, though the DS group had more episodes of fever, the duration and intensity of the fever was no different compared to the control group.

Of the 26 patients with DS who developed fever, 11 (42%) of them had blood cultures drawn and 6 had received a course of antibiotics waiting for the results of the culture. The control group had 11 patients who developed fever, of which 7 (64%) had blood cultures drawn and 4 received antibiotics. Of interest, however, is that the control group did receive statistically more total doses of antipyretics, per patient, in the first 72 hours than the DS group (Table 1).

The role of a fever in the operative and postoperative recovery phases in patients with DS is shown in Table 2. Overall, the patients who had fever were older than patients without fever. Yet, the remainder of operative and postoperative variables including their time on bypass as well as recovery from the ventilator, time in the ICU, WBC 24 hours postoperatively, and LOS were equal in patients with DS with and without fever. Furthermore, the surgical economic variables, including both cost and charge, were unchanged if fever was present during recovery in patients with DS.

OPEN IN VIEWER

Table 2.

Comparisons of Demographics, Surgical Variables, Outcomes, and Costs Between Patients With DS and Fever and Patients With DS Without Fever.^a

	Fever (n = 26)	No Fever (n = 19)	P Value
Age, days	146.5 ± 63.3	103.8 ± 45.8	.04
Weight at surgery, kg	5.4 ± 1.1	4.4 ± 0.9	.08
Female, n (%)	8 (31%)	11 (58%)	.11
CPB time, minutes	123.6 ± 72.2	125.3 ± 61.4	.94
Ventilator time, hours	63.7 ± 62.2	64.8 ± 60.2	.97
ICU LOS, days	5.7 ± 4.2	7.4 ± 3.9	.23
Total LOS, days	9.8 ± 8.4	11.3 ± 4.7	.54
Total hospital cost, US$	44,114 ± 27,362	47,655 ± 16,886	.67
Total hospital charge, US$	143,194 ± 80,511	151,436 ± 55,813	.75
WBC at 24 hours postoperative, k/µL	14.9 ± 3.4	13.0 ± 4.1	.13

Abbreviations: CPB, cardiopulmonary bypass; DS, Down syndrome; ICU, intensive care unit; LOS, length of stay; WBC, white blood cell count.

^aData are presented as mean ± standard deviation or n (%).

Discussion

This case–control study of infants with DS undergoing heart surgery is the first to demonstrate a higher prevalence of culture-negative fever in the first 72 hours postoperatively compared to age- and disease-matched children with no known chromosomal anomalies. Importantly, those patients with DS who do have postoperative fever do not have any difference in postoperative outcomes or costs in the short-term period. While increased rates of true infection have been clearly chronicled in patients with DS following cardiac surgery, ^2,3 higher prevalence of fever of unknown origin, or without a source, has not been previously described.

Cardiopulmonary bypass (CPB) stimulates a profound inflammatory process in children and adults. ⁷ Various pro-inflammatory cytokines and markers are significantly upregulated during CPB including interleukin 6 (IL-6), IL-8, and C-reactive protein (CRP). ^{8 –10} Tumor necrosis factor α (TNF-α) and IL-10, which have some anti-inflammatory properties, also demonstrate significant alterations after CPB. ^{11 –14} Furthermore, when evaluating a large portion of the proteome in infants following CPB, it is not surprising that major modifications are seen in many cytokines and biomarkers together. ^15,16 While total time on bypass did not differ among those patients with DS who had fever and patients with DS who did not, this suggests the presence of other inflammatory processes altered by CPB that were unmeasured as a result of the retrospective nature of this study.

While alterations in cytokines and pro-inflammatory markers following CPB are present in all infants, alterations have not been looked at specifically in patients with DS. This is an important concept as altered immunomodulation and abnormal cytokine response is known to be present in patients with DS when stimulated by other antigens and conditions. Baseline IL-6 levels, a prominent pro-inflammatory cytokine, have repeatedly been demonstrated to be higher in infants with DS compared to those without chromosomal anomalies, ^17,18 and this comparative effect holds true when patients are stimulated with antigen. ⁴ A similar effect has been demonstrated with IL-8, IL-10, TNF-α, interferon γ, and CRP levels. ^{17 –20} Furthermore, there is evidence that patients with DS also have significantly diminished expansion of B and T lymphocytes. ²¹ Given the altered immune response in patients with DS, it may not be surprising that these patients have issues with temperature regulation as part of the inflammatory response from CPB. While this study demonstrated an increased prevalence of fever in these patients, it is unknown what their individual cytokine response was during febrile events.

Infants with DS are well known to have issues with their thyroid and thyroid hormone regulation. Pediatric and adult patients have been shown to have low plasma concentrations of thyrotropin and other circulating thyroid hormones including Triiodothyronine (T3), free Thyroxine (fT4), and thyroglobulin following CPB. ^23,24 This combination suggests a possible role of thyroid hormone in temperature regulation in these patients, though actual thyroid function in this subset following CPB remains to be elucidated.

Total WBC count at the time of fever, if measured, was not appreciably different between the populations. While interesting to note, isolated total WBC count is felt to be a poor indicator of underlying infection or inflammation following congenital heart surgery. ²² Furthermore, the relatively few doses overall of antipyretics administered in both groups made further evaluation of this variable’s effect on the outcome challenging.

Similar to previous studies, ^2,3 our patients with DS demonstrated more morbidity in the form of longer ICU days, ventilator hours, and total hospital days compared to control infants undergoing similar operations. Despite the fact that fever did not seem to affect these variables, these findings continue to emphasize how major chromosomal anomalies alter recovery in pediatric patients undergoing heart surgery with CPB.

Conclusions

Among many other known comorbidities that can complicate postoperative management after cardiac surgery, patients with DS also have a higher incidence of culture-negative fever in the first 72 hours. However, the presence of fever in these patients does not affect their overall postoperative outcomes or costs. Future studies to assess cytokine levels, immune function, and thyroid function as possible etiologies of fever could provide additional insight to optimize the management of these patients.