Zeb, Amna MD^*; Rosenberg, Rebecca E. MD, MPH^*; Ahmed, ASM Nawshad Uddin FCPS^‡; Saha, Samir K. PhD^§; Chowdhury, MAK Azad FRCP^¶; Ahmed, Saifuddin PhD^†; Darmstadt, Gary L. MD, MS^*

Of the 4 million neonatal deaths each year, 99% occur in low and middle income countries, primarily from severe infections (36%), complications of preterm birth (28%), and birth asphyxia (23%).¹ Sclerema neonatorum is a disorder of adipose tissues, which predominantly affects preterm male neonates in the first week of life, manifesting as a hardening of skin and adipose tissues to such an extent that it hinders feeding and respiration, and usually culminates in death. It has been reported in both high and low income countries.^2–7

We describe neonates who developed sclerema neonatorum in the course of a clinical trial to evaluate the effects of topical treatment with skin barrier-enhancing emollients on prevention of sepsis and mortality in preterm neonates in Bangladesh,^8,9 identify risk factors and summarize implications for management of these infants.

MATERIALS AND METHODS

Study Participants

Newborn babies admitted to the Special Care Nursery at Dhaka Shishu Hospital, Bangladesh, from December 1998 to July 2003 were enrolled in the study. The hospital lacks a maternity ward but is the largest tertiary-level pediatric hospital in the country, with 8 incubators. Inclusion criteria were gestational age at birth ≤33 weeks (determined as described previously⁸) and age at presentation to the hospital ≤72 hours. Exclusion criteria included neonates with congenital anomalies or hydrops fetalis, those admitted to the hospital for a major surgical procedure or who had a clinically evident skin infection confirmed by surface culture, neonates with a generalized skin disease or a structural skin problem affecting >5% of the body surface area, and those judged by the admitting physician to be critically ill and likely to die within the first 48 hours of admission. After confirmation of eligibility by the doctor and administration of informed consent, infants were randomly assigned to 1 of 3 groups: treatment with sunflower seed oil or Aquaphor or the control group. Care of neonates in the hospital has been described previously,^8,9 and included feeding of breast milk whenever possible, but Kangaroo Mother Care had not yet been introduced.

Identification of Sclerema Neonatorum Cases

Sclerema neonatorum was diagnosed on clinical inspection based on daily examinations of the skin and subcutaneous tissues by 2 neonatologists. We did not use skin biopsy for the diagnosis but used the criteria of uniform hardening of skin and subcutaneous tissues to the extent that the skin could not be pitted nor picked up and pinched into a fold.^2–4

Data on Potential Risk Factors

Sociodemographic information, delivery, and postnatal history of the infants was derived by study physicians from the birth card for those born at a facility, as well as by self report from the family. Sociodemographic data included monthly income, area of residence, paternal education, consanguinity in marriage; and maternal age, education, parity, and body mass index. Delivery information consisted of place, birth attendant, season and type of delivery, and duration of rupture of membranes and of second stage of labor. Immediate postnatal data consisted of gestational age, sex, presence of cry at birth, strength of cry, presence of cyanosis, asphyxia and apnea at birth, and type of cord care. Data ascertained by physician examination at the time of admission to hospital consisted of infant weight, temperature, pulse, respiratory rate, heart rate, and mid upper arm circumference.

Daily clinical evaluations performed by physicians included assessment for 37 historic factors and clinical signs, as described previously,⁸ including lethargy, irritability, pallor, jaundice, dusky color/cyanosis, excess crying, poor feeding, vomiting, dehydration, abdominal distention, grunting, respiratory distress, apnea, signs of Sclerema (see above), petechiae, poor peripheral perfusion, seizure, decreased muscle tone, hepatomegaly, and splenomegaly. Infants were assessed at least 3 times daily.

Identification of Sepsis

Baseline blood cultures were done routinely at enrolment, within the first 48 hours of admission and before initiation of topical therapy, to establish whether bacteremia was present. Subsequent cultures of blood were obtained for suspected sepsis by applying the standard set of 37 historical factors and clinical criteria described previously.⁸ Methods of processing blood culture and cerebrospinal fluid specimens have also been described previously.^8,10,11

Statistical Analysis

Our primary outcome measure was development of sclerema neonatorum. We explored risk factors for the outcome by comparing patients with and without the condition, using one-way ANOVA for continuous variables and univariate logistic regression for categorical and binary variables. We then used multivariate logistic regression, adjusting for weight on admission, gestational age, and treatment group for those risk factors of potential importance (ie, P < 0.15) in the univariate analysis. For risk factor analysis of sepsis, we determined the incidence rate ratio of positive cultures among total hospital days for infants with and without sclerema neonatorum. All analyses were performed using Stata 9.2 (Stata Corporation, College Station, TX).

Approval for the Study

The study protocol was approved by the Johns Hopkins Committee on Human Research and the Ethical Review Committee of Dhaka Shishu Hospital. The trial was registered at clinicaltrials.gov (No. 98-04-21-03-2). Oral consent was obtained from a parent or guardian of every enrolled neonate.

RESULTS

Study Participants

A total of 4542 infants were screened for eligibility, of which 497 met enrolment criteria. Fifty-one infants developed sclerema neonatorum (10.3%). Baseline characteristics of the participants in the emollient trial have been described previously ^8,9; characteristics of the subset of patients with sclerema neonatorum are presented in Table 1 (see Supplemental Digital Content 1, http://links.lww.com/A883).

Course of the Disease

The mean age at diagnosis of sclerema neonatorum was 8.8 (SD: 6.1) days. Nearly half of the cases (n = 24 of 51, 47%) presented in the first 6 days of life and approximately two-thirds were diagnosed within the first 7 days of hospital stay (n = 35 of 51, 69%). Typically, the skin and subcutaneous hardening appeared first in the limbs, especially on the thighs and buttocks, and then became generalized. Common associated signs at diagnosis of cases were jaundice (n = 42 of 51, 82%), poor peripheral perfusion (n = 37 of 50, 74%), apnea (n = 24 of 50, 48%), and hyperthermia (n = 13 of 50, 26%). Case fatality in infants with sclerema neonatorum was 98% (50 of 51) as compared with 52% (233 of 446) of the controls. The risk of death in infants with the disease was 46.5 times higher than for those without it (P < 0.001, 95% CI: 6.37–339.81). The most common final diagnoses in infants with sclerema neonatorum were jaundice (n = 48 of 50, 96%), anemia (n = 42 of 50, 84%), and apnea (n = 39 of 50, 78%).

Risk Factors for Sclerema Neonatorum

Univariate Analysis

Among categorical variables, there was an important (P < 0.15) increase in risk of sclerema neonatorum with lower maternal education (OR: 1.87; 95% CI: 0.99–3.53; P = 0.057), maternal body mass index <18 (OR: 2.29; 95% CI: 0.81–6.48; P = 0.117), and presence of jaundice (OR: 2.05; 95% CI: 0.93–4.49; P = 0.075), poor feeding (OR: 2.9; 95% CI: 0.69–12.5; P = 0.144), and apnea (OR: 2.55; 95% CI: 0.98–6.6; P = 0.055) on admission (see Table 3, Supplemental Digital Content 2, http://links.lww.com/A884).

Multivariate Analysis

After controlling for gestational age, weight of neonate on admission to hospital, and treatment group; the factors associated with increased risk of developing sclerema neonatorum in multivariate analysis were lower maternal education (OR: 1.94; 95% CI: 1.02–3.69; P = 0.043), jaundice (OR: 2.82; 95% CI: 1.19–6.69; P = 0.018), and poor feeding (OR: 4.71; 95% CI: 1.02–21.74; P = 0.047) on admission to hospital.

Sepsis as a Risk Factor

A total of 497 baseline blood cultures were done at time of admission to the hospital and 196 blood cultures and 26 cultures of cerebrospinal fluid were done for suspected sepsis after admission to hospital. Thirty-nine percent (192 of 497) of neonates had at least 1 positive culture. Nearly half of infants with sclerema neonatorum developed a positive blood culture for a bacterial pathogen (n = 25 of 51, 49%) as compared with 33.4% (149 of 446) of controls. Infants with sclerema neonatorum had 1.81 times (95% CI: 1.16–2.73; P = 0.004) increased incidence of sepsis compared with the infants without the condition (6.04 vs. 3.33 episodes of sepsis per 100 days, respectively). Gram-negative organisms were the predominant causative agent of nosocomial sepsis in both cases and controls (Table 2).

TABLE 2

DISCUSSION

Lower maternal education, poor feeding and signs of jaundice on admission, and diagnosis of culture-proven sepsis were identified as risk factors for developing sclerema neonatorum. Poor feeding was also reported to be important in 2 case series in which most of the neonates with the condition (n = 12 of 17, 70.6%⁶ and n = 16 of 16, 100%⁷) had feeding difficulty either before or after the onset of the disease. Among signs present on admission, twice as many neonates who developed sclerema neonatorum (18%) were noted to have signs of jaundice as compared with the controls (9%). In other published case series, half of neonates with sclerema neonatorum developed jaundice (n = 5 of 9, 55.6% and n = 8 of 17, 47%).^2,6 The neonates who developed sclerema neonatorum in our study had a nearly 2-fold increased incidence of developing sepsis (approximately half of those with disease) as compared with neonates who did not have it (one-third). Other investigators have also reported that a substantial proportion of such neonates had culture-proven sepsis (n = 6 of 18, 33%³; n = 6 of 17, 35.3%⁶; and n = 7 of 16, 43.8%⁷).

The patients in our study population, in addition to being preterm, were acutely and severely ill on admission to the hospital, which serves as the primary pediatric tertiary referral center in Dhaka.¹² This could account for the high incidence of sclerema neonatorum in our study population. No other data on incidence of this entity in Bangladesh is available.

An important limitation of our study was that the sample was not representative of the whole population. We excluded cases with major congenital anomalies, which have been reported in case series of sclerema neonatorum.^3,7,13

The case fatality rate in our study was 98%, higher than reported in previously published case series: 61% and 78% from the United States, 79% from the United Kingdom and United States, 82% from India, and 84% and 88% from Africa.^2–4,6,7,14 We found the risk of death to be 46.5 times higher for neonates with sclerema neonatorum compared with controls. This could be linked to the high rate of sepsis in these neonates, as neonatal sepsis is fatal when associated with the condition.^4,15,16 This makes prompt diagnosis and treatment of this disease, including underlying sepsis, imperative to save the lives of these neonates, and findings from microbiological analysis suggests that initial empirical antibiotic treatment must include effective Gram-negative coverage.

Among the intrapartum and postnatal risk factors, prolonged rupture of membranes, cyanosis at birth, male sex, and hypothermia did not prove significant in our study. These have been reported as factors associated with sclerema neonatorum in previous series of case reports.^2–4,6,7,13 In a matched case control study, the only other published study of risks factors for sclerema neonatorum that we are aware of, Ji et al¹⁷ reported gestational age <37 weeks and birth weight <2500 g as risk factors. Within our preterm study population, we attempted to control for confounding by gestational age and birth weight when examining other potential risk factors. In contrast to other studies which suggested that respiratory problems (eg, pneumonia, respiratory distress at or shortly after birth with evidence of hyaline membrane disease on autopsy,^3,4,7 and birth asphyxia¹⁷) were associated with sclerema neonatorum, we did not find any association of respiratory signs with development of the disease.

At the time of diagnosis of sclerema neonatorum, poor peripheral perfusion, a sign of low peripheral vascular pressure, was noted in three-fourths of cases. This is in accordance with the theory proposed by Hughes and Hammond⁴ that affected neonates have a high ratio of saturated fat in their subcutaneous tissues, which hardens due to a decrease in body temperature as a consequence of the circulatory collapse of shock. Since poor peripheral perfusion was not shown to be a risk factor for preterm neonates to develop sclerema neonatorum, other theories of pathogenesis must be considered.^2–5,18,19 Milunsky and Levin¹³ have also argued against the theory of Hughes and Hammond by stating that exposure to cold causes the blood circulation to become sluggish, causing peripheral vasoconstriction and tissue anoxia. Intracellular and intravascular fluid thus moves into the interstitial spaces, rendering the hardened feel to the skin and subcutaneous tissues that characterizes the condition. Our study, however, did not show hypothermia to be associated with sclerema neonatorum either. Thus, pathogenesis of sclerema neonatorum remains poorly understood.

REFERENCES

Supplemental Digital Content

• INF_28_5_DARMSTADT_201396_SDC1.doc; [Word] (62 KB)
• INF_28_5_DARMSTADT_201396_SDC2.doc; [Word] (135 KB)