Canadian Paediatric Surveillance Program - 2003 Results

Lap-belt syndrome

(September 2003 to August 2005)

Highlights

• Seat belts were designed to save lives, but worn incorrectly they can cause important abdominal and lumbar spine injuries.
  
• Three cases of lap-belt syndrome were reported in the first four months, including one child with spinal cord injury.
  
• Media, public health and general interest in the study has been high.
  

Background

The use of seat belts has clearly reduced fatalities and severity of injuries in motor vehicle crashes. With the increasing use of seat belts over the last decades, a new association of injuries has emerged among adults and children involved in motor vehicle crashes. The 'seat-belt syndrome' was first described by Garrett and Braunstein in 1962 and refers to injuries to the intestinal viscera and to the lumbar spine associated with seat-belt restraints. Children are especially vulnerable to these injuries as their intra-abdominal organs are less protected by the thorax and pelvis, they have a lower centre of gravity and their iliac crests are less developed than those of adults allowing the belt to ride up over the abdomen. To date, there have been very few paediatric studies on the incidence of seat-belt syndrome. In fact, most current knowledge comes from case reports or studies done in limited regional areas. In these studies the number of cases was relatively low, ranging from ten to 50 cases over years.

Objectives

1. Obtain epidemiologic data on the incidence and pattern of injuries encountered in the seat-belt syndrome.
2. Identify at risk age groups.
3. Supply data that will help develop new strategies in order to adequately protect children in motor vehicles.
4. Promote education and awareness of this rare disease among health-care professionals.

Case definition

Any child up to and including 18 years of age restrained in a motor vehicle at the time of a crash, with either an abdominal injury, as determined by operation or CT-scan, or thoraco-lumbar spine injuries with or without spinal cord injuries.

Results and discussion

The study on lap-belt associated injuries started in September 2003. In the first four months of the study, Canadian paediatricians reported eight cases of lap-belt syndrome. Of these, three were confirmed, two were duplicates, one was a discard, and two had injuries probably related to their seat belts, but which did not meet the case definition (one child was in a booster seat and had a cervical spine fracture, while the other presented with a brachial plexus injury). Of the three confirmed cases, one presented with abdominal abrasions and a splenic contusion, another also had an abdominal abrasion and a jejunal infarct that needed small bowel resection, while the last presented with the full spectrum of seat-belt associated injuries, including an intestinal perforation that needed resection and an L2-L3 lumbar spine fracture with spinal cord injury.

Lap-belt syndrome is an important health problem in the Canadian paediatric population. It is associated with improper use of the lap belt only. Most encouragingly, even though cases have just started to be reported, interest from different media and public health organizations has been high. Issues pertaining to child safety in motor vehicles and the use of booster seats have been discussed not only with radio and newspaper journalists but also with the Ontario Ministry of Transport. The general public is already being informed and some provinces are taking legislative measures to ensure that children are adequately protected in motor vehicles. Clearly the most important issue of this study is to demonstrate the need to review restraints in motor vehicles. It is most satisfying to see that this objective is already being fulfilled.

Principal investigator

Claude Cyr, MD, Centre hospitalier universitaire de Sherbrooke, 3001 12e Ave N, Sherbrooke QC J1H 5N4; tel.: 819-346-1110, ext. 14634; fax: 819-564-5398; e-mail: claude.cyr@courrier.usherb.ca

Co-investigators

Claude Lemoine, MD, Département de pédiatrie, Centre hospitalier universitaire de Sherbrooke

Miriam Santschi, MD, Département de pédiatrie, Centre hospitalier universitaire de Sherbrooke

Necrotizing fasciitis

(September 2001 to August 2003) Final report

Highlights

• Sixteen of the 26 (61%) GAS-related necrotizing fasciitis cases were varicella-associated.
• Implementation of universal varicella immunization would decrease NF incidence.
• Excruciating local pain out of proportion to skin presentation warrants urgent attention.
• Mortality was low but morbidity was high.
• All cases required surgery.

Background

In 1999, the Canadian Paediatric Society issued a statement on the state of knowledge and management of children and close contacts of persons with all-invasive group A ß-hemolytic streptococcal (GABHS) infections. In that statement, it was noted that there was no national data for necrotizing fasciitis (NF) in Canada. The current study started in September 2001 and concluded in August 2003 using the Canadian Paediatric Surveillance Program (CPSP) to establish actual national rates and epidemiology of NF.

Objectives

To define the epidemiology, management and outcome of NF in Canadian children.

Case definitions

NF is a deep-seated infection of the subcutaneous tissue that results in progressive destruction of fascia and fat. In general, NF is classified into two types.

1. Type I NF refers to mixed infections involving anaerobes (most commonly Bacteroides and Peptostreptococcis spp) and one or more facultative anaerobes, such as streptococci (non-GABHS), and members of the Enterobacteriaceae (e.g., Escherichia coli, Enterobacter, Klebsiella, Proteus).
2. Type II NF refers to that caused by invasive GABHS.

Results

Thirty-seven cases of NF were reported (26 group A streptococcus [GAS] related) across Canada with a mean age of 6.3 ± 4.9 years. Cases were reported from all areas of Canada west of the Maritimes. The annual incidence per million population peaked during the first year of life: age less than one year, 8 cases; 1-9 years, 3.4 cases; and 10-18 years, 1.1 cases, (P<0.001). Males under the age of one had the highest burden of disease with 12 cases per million population versus 3.2 cases per million for females (P<0.0001). In contrast, rates were similar between males and females for older age groups, (1-9 years: males, 3.2 cases, and females, 3.7 cases; 10-18 years: males, 1.0, and females, 1.25 cases, P=NS). Seventeen (46%) of all cases occurred within one month of the child developing varicella. Sixteen (61%) of the GAS related type II NF cases were varicella-associated versus one (9.1%) of the non-GAS type I related cases (P<0.002). Eight (21.6%) patients had an underlying chronic medical condition (2/26 type II related versus 6/11 type I related cases, P<0.01). For the 26 cases of type II related NF, the most common presenting symptoms were: localized pain (25, 96.2%), fever (10, 38.5%), vomiting (9, 34.6%) and chills (8, 30.8%). Localized pain was also reported uniformly in type I cases (11, 100%) along with chills (6, 54.5%) and fever (3, 27.1%). Eleven (42%) of the type II patients gave a history of having taken non-steroidal anti-inflammatory drugs (NSAIDS) within a week of admission. One type II patient had been in contact with another patient with invasive GAS disease. Two (18.2%) of the type I cases and none of the type II cases were nosocomial. Eighteen (48.6%) cases involved the lower extremities and/or groin area; nine (24.3%) the upper extremities (no type I cases involved upper extremities); nine (24.3%) the head, neck and chest areas; and two (5.4%) involved the abdomen (some patients had more than one site involved). All patients received at least one surgical procedure including fasciotomy/fasciectomy, debridement, amputation or skin graft. All type II patients received clindamycin in combination with either penicillin (20 patients) or other beta-lactam antibiotic (6 patients). Overall, seven of the 37 patients (5/26 type II and 2/7 type I) were noted to have received IVIG, while another 14 (10/26, 38.5% type II and 4/11, 36.4% type I) patients received unspecified blood products. The mean length of stay was 16.8 ± 15.8 days. Two patients died (0/26 type II and 2/11, 18.2% type I).

Discussion

NF occurred throughout Canada, with the peak incidence among males younger than one year of age. The majority of type II NF cases were associated with recent varicella infection, whereas type I cases were more likely to be associated with underlying disease or be nosocomial. The lower extremities and groin areas were the most commonly involved parts of the body. Although outcomes with regard to mortality were excellent, there was substantial morbidity in the form of surgical procedures and prolonged hospital stay. This improved mortality may be related to one or more of the following: universal surgical intervention, early use of antibiotics and use of IVIG.

Conclusions

Surveillance through the CPSP has led to a better understanding of the epidemiology of paediatric NF in Canada. Pain out of keeping with clinical findings is an early symptom for clinicians. Type I and type II NF differ in terms of predisposing risk factors, relationship to varicella and possibly prognosis. This morbidity associated with type II NF supports the need for a universal varicella immunization program in Canada. Even though the largest at-risk population is under one year of age, "herd immunity" would be expected to reduce NF rates in this population.

Principal investigator

H. Dele Davies, MD, Pediatrics and Human Development, Michigan State University, B240 Life Sciences Building, East Lansing MI 48824-1317, USA; tel.: 517-353-5042; fax: 517-353-8464; e-mail: daviesde@msu.edu

---

Neonatal herpes simplex virus infection

(October 2000 to September 2003) Final report

Highlights

• Nine children died, resulting in a case fatality rate of 15.5%.
• Eight of 20 (40%) women were unaware of a history of HSV infection prior to delivery.
• The majority of typed cases (62.5%) were HSV-1, which has implications for vaccine and drug development.

Background

Herpes simplex virus (HSV) infections pose a public health concern especially since a high proportion of these infections are unrecognized. The most serious direct consequence of genital HSV infection is the perinatal transmission from mother to infant. With only limited data available, it is not possible to accurately determine the prevalence, incidence and trends of neonatal herpes infection in Canada. Data collection is essential to better understand the epidemiology of neonatal herpes and to monitor the trends in Canada. Canadian data on morbidity, mortality and on mother and infant risk determinants will allow comparison of neonatal herpes infection rates with other countries. The pre-vaccine baseline data that will be collected through this CPSP study before an HSV vaccine becomes available will be used to promote prevention and control program strategies, to further research, and to estimate the burden of illness in Canada.

Objectives

1. To estimate the incidence rate of neonatal herpes infections (HSV-1 and HSV-2) for the years 2000 to 2003 per 100,000 live births in Canada.
2. To determine the proportion of HSV-infected infants with localized and disseminated disease.
3. To identify risk determinants in mothers and their maternal HSV status prior to delivery.
4. To analyze trends of cases reported over a minimum period of three years, by age, sex and province.
5. To document the morbidity/mortality of neonatal infections for the years 2001 to 2006 through a cohort study of infants identified in each of the first three years of the neonatal herpes surveillance project.

Case definition

For the purpose of this study, the neonatal period was extended to 60 days of life so that late diagnoses were not missed, thereby optimizing the capacity to identify the maximum number of cases. All cases were laboratory-confirmed by at least one of the following tests:

1. Culture
2. HSV IgM
3. Polymerase chain reaction (PCR) in an infant equal to or less than two months (60 days), born between October 1, 2000 and September 30, 2003, who demonstrates one of the following:

• localized infection involving the skin, eyes or mouth,
• disseminated infection:
  a) to central nervous system (CNS)
  b) to organs other than CNS.
  

Results and discussion

The status of all cases reported during the three-year study period was resolved as outlined below in Table 21.

A total of 58 cases of neonatal herpes simplex infection were confirmed (5.9 per 100,000 live births), nine of which were fatal. The majority of cases were reported from Central Canada (60.4%), with the remainder from Western Canada (6.9%), the Prairies (25.8%) and Atlantic Canada (6.9%). No cases were reported from Northern Canada. For reporting purposes, the year of diagnoses of a positive laboratory test for HSV was used. The high number of duplicate reports of individual cases (44/122) is indicative of physicians' buy-in to this comprehensive voluntary mechanism of case ascertainment.

The overall demographic and health profile of the 58 confirmed neonatal HSV ases is summarized in Tables 22 and 23, for mothers and infants respectively. Table 22 outlines that the majority of mothers were Caucasian, with a mean age of 27, and that most deliveries were vaginal (75.9%) versus Caesarian (24.1%). Importantly, of the 20 women with available information, 12 had a history of HSV infection. Out of 58 cases, only one mother had intrapartum genital HSV lesions.

Approximately half of the neonatal cases were diagnosed by 11 days old, and 51.7% of these cases were female (Table 23). Of the 48 neonates with known HSV type, 18 (37.5%) had HSV-2 and 30 (62.5%) were infected with HSV-1. Over a quarter of cases (27.6%) were delivered prematurely, although the median Apgar score was nine. The majority of infections were localized (63.8%), with almost 90% of these cases localized to the skin.

TABLE 21 Neonatal herpes simplex virus infection reported cases
Status	2000	2001	2002	2003	Total
Confirmed NHSV*	4	24	17	13	58
Cases pending review	0	0	0	0	0
Did not meet entry criteria†	4	8	7	1	20
Duplicates	2	20	18	4	44
Total	10	52	42	18	122

TABLE 22 Demographic and health profile of the mothers of neonatal HSV cases diagnosed from October 2000 through September 2003 (n=58)
Mean age (years)	27
Ethnicity:
• Caucasian	74.1%
• Aboriginal	10.3%
• Black	5.2%
• Other	10.4%
Delivery type:
• Caesarian	24.1%
• Vaginal	75.9%
History of HSV infection prior to delivery*	60.0%
Presence of intrapartum genital HSV lesions	1.7%
HIV infected	0%
* Information available to paediatricians for 20 mothers

TABLE 23 Demographic and health profile of the neonates diagnosed with HSV from October 2000 through September 2003 (n=58)
Female	51.7%
Mean gestational age (weeks)	37.8
Preterm birth (<37 weeks)	27.6%
Mean birth weight (grams)	2906
Median Apgar score at 5 minutes	9
Median age at laboratory
diagnosis (range)	11.8 (0-45) days
HSV type*:
• HSV-1	62.5%
• HSV-2	37.5%
Classification of HSV infection†:
• Localized	63.8%
• Disseminated	34.5%
* n=48 as ten surveys were missing information on virus type. † n=57 as one survey was missing information on nature of infection.

Table 24 outlines a comparison of type of infection and age of diagnosis by infant virus type. For the 47 infants with known virus type and type of infection, localized infections were more likely to be caused by HSV-1, while disseminated infections were more likely to be caused by HSV-2 (p=0.017). For the 46 infants with known virus type and age of diagnosis, 19 of 28 HSV-1 infections (67.9%) were diagnosed after the first week of life, while eight of 18 HSV-2 infections (44.4%) were diagnosed in this time frame (p=0.116).

Of the 56 neonates with known treatment information, 98.2% received treatment, all with acyclovir. In the first two months of life, 14 cases (24.1%) displayed obvious sequelae of infection, including seizures (9), encephalitis (10), microcephaly (2), and blindness (1). The overall case fatality rate was 15.5%, with death more likely to occur among disseminated than localized cases (36.8% vs. 2.8%, p<0.001). Of the nine fatal cases, deaths occurred in infants with the following: known disseminated HSV infection (7), localized infection only, negative CSF results on PCR, and respiratory failure secondary to severe CNS insult listed as cause of death (1), and no information on the nature of the infant's infection available with an autopsy pending (1). All nine infants died within 24 days of birth, with seven cases having dissemination to the CNS, four with dissemination to the liver, and three with dissemination to the lungs. For the eight fatal cases with known virus type, six were typed HSV-2 while only two were typed HSV-1 (p=0.016).

TABLE 24 Comparison of type of infection and age of diagnosis by virus type
	HSV-1	HSV-2	p-value
Type of infection (n=47)
• Disseminated	5	9	0.017
• Localized	24	9
Age of diagnosis (n=46)
• 0-7 days	9	10	0.166
• >7 days	19	8

Conclusions

Based on 58 confirmed cases from October 2000 through September 2003, the reported neonatal herpes incidence rate in Canada was 5.9 per 100,000 live births. This represents a rate that is closer to that reported by the United Kingdom (two per 100,000 live births) than by the United States (20-50 per 100,000 live births). Over a third of these infections were disseminated cases, with an overall case fatality rate of 15.5%.

The results of this surveillance project have a number of public health implications. Not only is the predominance of neonatal herpes cases attributed to HSV-1 consistent with findings from a growing body of research that indicates an increase in the proportion of genital herpes cases attributed to HSV-1, but the data also shows that herpes vaccine and drug development research for genital herpes needs to account for HSV-1 and HSV-2 genital infections, and demonstrate the effectiveness against both. In addition, study results outline the challenge of preventing neonatal HSV infection and implications for prenatal screening, given that eight out of 20 mothers were unaware of their herpes infection prior to delivery, and obvious genital lesions were present in only one out of 58 cases during the intrapartum period.

As per the project's fifth objective, surviving cases will continue to be followed annually for three years in order to evaluate the overall health and developmental consequences of neonatal HSV infection.

Principal investigator

Tom Wong, MD, Community Acquired Infections Division, Centre for Infectious Disease Prevention and Control, Health Canada, Room 3444, Building #6, Tunney's Pasture, Ottawa ON K1A 0L2; tel.: 613-957-1080; fax: 613-957-0381; e-mail: Tom_Wong@hc-sc.gc.ca

Co-investigators

Sandra Burton, BSc, Community Acquired Infections Division, Health Canada

Joanne Embree, MD, University of Manitoba

Rhonda Kropp, BSN MPH, Community Acquired Infections Division, Health Canada

Marc Steben, MD, Institut national de santé publique du Québec, Montréal

---

Neonatal hyperbilirubinemia -severe

(July 2002 to June 2004)

Highlights

• In 2003, 143 cases of newborns affected with severe hyperbilirubinemia were confirmed.
• No etiology was reported in over 70% of cases.
• At presentation, it is essential to perform all laboratory investigations to confirm the etiology.

Background

Even though the occurrence of severe neonatal hyperbilirubinemia and bilirubin encephalopathy is very rare, it can be associated with significant morbidity. Bilirubin encephalopathy is a condition that is unfamiliar to most paediatricians practicing today. In the 1940s and 1950s, bilirubin encephalopathy was a common complication of hyperbilirubinemia associated with Rhesus (Rh) disease and occasionally with ABO hemolytic disease. With the introduction of exchange transfusions and Rh immunoglobulin, a reduction in the occurrence of bilirubin encephalopathy was noted. Also, better antenatal monitoring and availability of intrauterine blood transfusion has eliminated most cases of erythroblastosis fetalis secondary to Rh disease. Phototherapy has drastically reduced the need for exchange transfusions. Despite this, in the last several years, reports of bilirubin encephalopathy associated with extremely high serum bilirubin levels have increased (Penn et al., 1994, MacDonald et al., 1995, Maisels et al., 1995). In most cases, term infants appeared to be healthy and breast-fed with no evidence of obvious hemolytic disease (Rh disease or other antibody-related hemolysis).

Based on epidemiological studies, a number of risk factors have been found to be associated with severe hyperbilirubinemia in the newborn. These include jaundice presenting in the first 24 hours, jaundice noted at discharge from the hospital, previous sibling with jaundice, gestational age between 35 and 38 weeks, breast-feeding and infant bruising and cephalhematoma (Dennery et al., 2001, Newman et al., 2000). Additional risk factors identified by laboratory investigations include Rh and ABO incompatibility and glucose 6 phosphate dehydrogenase (G6PD) deficiency.

The frequency of severe neonatal hyperbilirubinemia during the current era has not been well documented. Attempts to better quantify its frequency, etiologies and associated risk factors in Canada would be of value prior to identifying strategies for risk reduction. Information obtained from a screening program for the detection of G6PD deficiency or routine determination of blood group and Coombs' analysis on cord blood may help to achieve risk reduction.

Objectives

The main study objective is to obtain epidemiological data on the incidence of severe neonatal hyperbilirubinemia and bilirubin encephalopathy, the burden of illness with regard to medical treatment (phototherapy, transfusions and exchange transfusions), and the neurodevelopmental outcome. Attempts will be made to identify the timing of presentation of jaundice, as well as the etiology and associated triggering or risk factors. This information will help in the development of prevention strategies (G6PD deficiency screening program, cord blood group and Coombs' test and educational programs).

Case definition

Term infants 60 days of age or less with unconjugated hyperbilirubinemia who have had either:

1. Peak serum total bilirubin >425 µmol/L or
2. Neonatal exchange transfusion

Exclusion criteria
Infants who have had exchange transfusion for well-documented Rh isoimmunization disease or are less than 36 weeks gestational age.

Results

Of the 178 cases of severe neonatal hyperbilirubinemia that were reported in 2003, 143 met the criteria for inclusion, eight are still pending review, 19 were duplicates and a further eight cases were discarded. The cause of severe hyperbilirubinemia was identified in less than a third of the cases (42/143). The etiologies included: ABO incompatibility (28), G6PD deficiency (7), other blood group incompatibility (2), urinary tract infections (2), spherocytosis (1), pyuvate kinase deficiency (1) and congenital hypothyroidism (1). The average peak bilirubin reported was 477 µmol/L with a range of 137-773 µmol/L. The infant with a bilirubin level of 137 µmol/L required an early exchange transfusion. Of the 143 confirmed cases, 126 neonates required phototherapy, 33 had an exchange transfusion while seven had transfusions of packed RBC's. Non-standard treatments included IVIG, albumin and phenobarbital.

TABLE 25 Severe neonatal hyperbilirubinemia reported cases
	Reported	Confirmed	Duplicates	Discarded	Detailed report form not returned
2002	91	60	10	17	4
2003	178	143	19	8	8
Total	269	203	29	25	12

Conclusions

Severe neonatal hyperbilirubinemia continues to occur in term neonates. In a significant proportion of reported cases, the underlying etiology could not be identified. This is partly attributable to incomplete evaluation of these infants at the time of admission. This finding highlights the importance of a complete hematological workup at the time of presentation, including a CBC, peripheral smear, screen for maternal and infant blood group, Coombs' testing and a G6PD screen. ABO incompatibility and G6PD deficiency were identified as important etiological agents in screened infants. These findings suggest not only that consideration should be given to implementing routine neonatal screening, but also that increased awareness and appropriate feedback to practicing physicians responsible for care of such neonates are needed.

Severe neonatal hyperbilirubinemia and its associated long-term neurological sequelae is a potentially preventable disorder. Several of the cases in this study were neurologically abnormal at the time of presentation. Although this study wasn't designed to assess the incidence of long-term neurological disease, such as kernicterus, a subsequent study addressing questions pertaining to the prevalence of kernicterus and its associated morbidity would be invaluable.

Principal investigator

Michael Sgro, MD, University of Toronto, Department of Paediatrics, Saint Michael's Hospital, 30 Bond St, Toronto ON M5B 1W8; tel.: 416-864-6060, ext 6560; fax: 416-864-5344; e-mail: sgrom@smh.toronto.on.ca

Co-investigator

Vibhuti Shah, MD, University of Toronto, Department of Paediatrics, Mount Sinai Hospital

Douglas M. Campbell, MSc, MD, University of Toronto, Department of Paediatrics, Saint Michael's Hospital

[ Previous ] [ Next ]