David A. Chiasson

The Diagnostic Value of the Diatom Test for Drowning, I. Utility: A Retrospective Analysis of 771 Cases of Drowning in Ontario, Canada

The utility and validity of the diatom test for drowning was studied using a retrospective analysis of 771 cases of drowning mostly from Ontario, Canada, over the period 1977 to 1993. In this article (part one), the utility of the test was assessed using an analysis of test outcomes. In the companion article (part two), the validity of the test was assessed by analyzing the relationship between test outcome and characteristics of diatoms in the bone marrow and samples of putative drowning medium. In the present study, freshwater drownings accounted for 738 of the cases and 33 cases were drownings in bathtubs, pools, or toilets. Diatoms were recovered from the femoral bone marrow of 205 cases (28%) of freshwater drowning and four cases (12%) of domestic water drowning. There was a monthly variation in the frequency of positive test outcomes that could not be explained by seasonal differences in the total number of drownings. However, the monthly variation was strongly correlated with the periodic cycle of diatom blooms that occurs in freshwater. Positive diatom tests were characterized by a limited number of distinctive diatom species per case, and a restricted quantity and size range of diatom frustules. These results indicate that the diatom test for drowning will identify approximately one in three victims of freshwater drowning and may be useful in the assessment of deaths occurring in bathtubs. The correlation of the outcome of the diatom test for drowning with diatom blooms provides further evidence for the reliability of the test.

Treatment of Campylobacter pylori-associated antral gastritis in children with bismuth subsalicylate and ampicillin

We evaluated the efficacy of medical therapy, consisting of liquid bismuth subsalicylate prescribed either in combination with oral ampicillin (n = 15) or alone (n = 1), in the clearing of bacterial colonization on the antrum in 16 children with Campylobacter pylori-associated antral gastritis. We also examined the effects of medical treatment on altering the severity of associated antral inflammation. Eight patients had upper gastrointestinal tract hemorrhage, two had acute gastric outlet obstruction, and 10 had symptoms of episodic epigastric abdominal pain. Duodenal ulcers were demonstrated in 10 of the 16 patients; in the other six, C. pylori-associated antral gastritis was documented without evidence of acute peptic ulceration. Seven days after a 6-week course of medical therapy, repeat upper endoscopy plus mucosal biopsy specimens showed that C. pylori colonization of the antrum had cleared in 12 of the 16 (75%) patients. Inflammation in the antrum improved in all patients in whom colonization by C. pylori was eradicated. In contrast, in the four with persistent colonization of the antrum, the severity of antral gastritis had not improved (p less than 0.01). Clinical symptoms improved in 9 of 12 patients in whom C. pylori colonization was no longer present, whereas subjective symptoms were unaffected in those with continued bacterial colonization of the antrum (p less than 0.05). We conclude that oral bismuth subsalicylate, in conjunction with ampicillin, can eradicate C. pylori colonization of the antrum, and that clearing of C. pylori is correlated with an improvement in the associated antral gastritis and clinical symptoms. These findings provide additional support for the hypothesis that these gastric organisms could play an etiologic role in primary antral gastritis and peptic ulcer disease.

Fracture of the hyoid bone in strangulation: comparison of fractured and unfractured hyoids from victims of strangulation.

The hyoid is the U-shaped bone of the neck that is fractured in one-third of all homicides by strangulation. On this basis, postmortem detection of hyoid fracture is relevant to the diagnosis of strangulation. However, since many cases lack a hyoid fracture, the absence of this finding does not exclude strangulation as a cause of death. The reasons why some hyoids fracture and others do not may relate to the nature and magnitude of force applied to the neck, age of the victim, nature of the instrument (ligature or hands) used to strangle, and intrinsic anatomic features of the hyoid bone. We compared the case profiles and xeroradiographic appearance of the hyoids of 20 victims of homicidal strangulation with and without hyoid fracture (n = 10, each). The fractured hyoids occurred in older victims of strangulation (39 +/- 14 years) when compared to the victims with unfractured hyoids (30 +/- 10 years). The age-dependency of hyoid fracture correlated with the degree of ossification or fusion of the hyoid synchondroses. The hyoid was fused in older victims of strangulation (41 +/- 12 years) whereas the unfused hyoids were found in the younger victims (28 +/- 10 years). In addition, the hyoid bone was ossified or fused in 70% of all fractured hyoids, but, only 30% of the unfractured hyoids were fused. The shape of the hyoid bone was also found to differentiate fractured and unfractured hyoids. Fractured hyoids were longer in the anterior-posterior plane and were more steeply sloping when compared with unfractured hyoids. These data indicate that hyoids of strangulation victims, with and without fracture, are distinguished by various indices of shape and rigidity. On this basis, it may be possible to explain why some victims of strangulation do not have fractured hyoid bones.

Value of postmortem thoracic CT over radiography in imaging of pediatric rib fractures

BackgroundStudies have reported that thoracic CT may provide greater sensitivity compared with radiography in detection of pediatric rib fractures and fracture healing. The additional sensitivity afforded by thoracic CT may have medicolegal implications where abuse is suspected.ObjectiveTo determine the additional value of postmortem thoracic CT compared with radiography in detecting pediatric rib fractures, and fracture healing, using autopsy findings as a gold standard.Materials and methodsWe retrospectively reviewed 56 coroner’s cases with postmortem radiography and CT thoracic survey. All studies underwent primary interpretation by one or two radiologists. The study radiologist independently reviewed all images from 13 patients with positive findings on radiography, CT or autopsy. Sensitivity and specificity between observers and imaging modalities were compared.ResultsPrimary interpretation: Fractures were recognized on radiography in 5/12 patients who had fractures found at autopsy, and on CT in 8/12 patients. In total, 29% (24/83) of fractures were reported on radiography, and 51% (52/101) of fractures were reported on CT. Study radiologist: Fractures were recognized on radiography in 7/12 patients who had fractures found at autopsy, and on CT in 11/12 patients. In total, 46% (38/83) of fractures were reported on radiography, and 85% (86/101) of fractures were reported on CT.ConclusionPostmortem thoracic CT provides greater sensitivity than radiography in detecting pediatric rib fractures, most notably in anterior and posterior fractures. However, the degree of improvement in sensitivity provided by CT might depend on observer experience.

The Location of Hyoid Fractures in Strangulation Revealed by Xeroradiography

The location of 19 fractures of 15 hyoid bones was studied from 13 cases of manual strangulation and two cases of hanging. The precise location of the fractures were determined by xeroradiography of isolated hyoid bones. To classify the location of the fractures, the greater cornu was divided into anterior, middle, and posterior thirds and the frequency of fracture occurrence in these locations was determined. In addition, the angle of curvature of the greater cornu was determined at the fracture site to ascertain if specific points along the greater cornu were more susceptible to fracture. Fractures were found with equal frequency in the posterior and middle thirds of the hyoid greater cornu (9/19, 47% for both posterior and middle) but were rare in the anterior portion (1/19, 5%). However, all fractures occurred between 30 degrees and 60 degrees of curvature of the hyoid and most fractures (approximately 60%) were found at approximately 50 degrees. These results indicate that fractures of the hyoid occur at vulnerable angles of curvature of the hyoid bone which, due to anatomic variation in hyoid shape, do not necessarily occur at specific segments of the hyoid bone. On this basis, although the presence of the hyoid fracture in strangulation is determined by the rigidity of the bone, the specific location is determined by the shape of the greater cornu.

Fatal child abuse-maltreatment syndrome. A retrospective study in Ontario, Canada, 1990-1995.

Child death due to repeated episodes of physical assault or neglect has been termed the child abuse-maltreatment syndrome (CAMS). We characterized the injuries in a series of fatally abused or maltreated child to delineate objective diagnostic criteria for the CAMS for use by clinicians and pathologists. All deaths (age <17 years) investigated by the Office of the Chief Coroner for Ontario, Canada during the time period 1990-1995 were reviewed. Cases of CAMS were defined as death due to lethal recent injury or malnutrition in the presence of significant old (healing or healed) injuries indicative of repeated episode of inflicted trauma. The nature and frequency of the various injuries was determined. The frequency of the shaken baby syndrome, and the types and frequency of ano-genital injuries were also studied. Twenty-one cases of fatal CAMS were found in the study period. Most cases had significant recent head injury with intra-cranial hemorrhage (71%). Other significant recent injuries commonly observed included blunt injuries of the skin and soft tissues (67%), blunt abdominal trauma with visceral injuries (14%), and fractures (18%). Eight cases (38%) fulfilled accepted criteria for the shaken baby syndrome. Many children with fatal head injuries had evidence of older head trauma (38% of all cases). A significant minority of cases had evidence of malnutrition due to neglect (10%) or ongoing ano-genital injuries (10%). Most cases of child homicides due to repeated episodes of abuse or maltreatment involve head trauma including shaken baby syndrome. Fractures of long bone and ribs, the classical markers of child abuse, were relatively infrequent compared with head injury. A proportion of cases had ano-genital injuries due to repeated sexual abuse or punitive maltreatment. All clinicians and pathologists must recognize the wide spectrum of injuries in child abuse to ultimate protect the victim or other children in an at-risk situation.

Increased incidence of CPR-related rib fractures in infants--is it related to changes in CPR technique?

OBJECTIVE A recent increase in the number of infants presenting at autopsy with rib fractures associated with cardio-pulmonary resuscitation (CPR) precipitated a study to determine whether such a phenomenon was related to recent revision of paediatric resuscitation guidelines. METHODS We conducted a review of autopsy reports from 1997 to 2008 on 571 infants who had CPR performed prior to death. RESULTS Analysis of the study population revealed CPR-related rib fractures in 19 infants (3.3%), 14 of whom died in the 2006-2008 period. The difference in annual frequency of CPR-related fractures between the periods before and after revision of paediatric CPR guidelines was statistically highly significant. CONCLUSIONS The findings indicate that CPR-associated rib fractures have become more frequent in infants since changes in CPR techniques were introduced in 2005. This has important implications for both clinicians and pathologists in their assessment of rib fractures in this patient population.

Pediatric drowning: a 20-year review of autopsied cases: III. Bathtub drownings.

Bathtub drownings are a significant cause of mortality in the pediatric population. Infants and preambulatory children are disproportionately affected, and several studies have suggested that preventative campaigns have been ineffective in the prevention of such deaths. To obtain a better understanding of the factors associated with bathtub drownings, a retrospective review of autopsy records over a 20-year period (1984–2003) was performed. Eighteen consecutive cases of bathtub drownings were identified in 8 males and 10 females (ratio, 0.8; P = 0.6374). The age ranged from 6 months to 70 months (mean, 17 months; median, 11 months), and most cases occurred in infants aged 12 months or less (72%). Males tended to be older than females (mean, 23 months versus 11 months; P = 0.1889). Associated factors included inadequate adult supervision (89%), cobathing (39%), the use of infant bath seats (17%), and coexistent medical disorders predisposing the infant or child to the drowning episode (17%). The pathologic findings included a frothy exudate (28%), pleural effusion (28%), and increased lung weight (61%). All toxicologic samples submitted for analysis were negative. The present study highlights the factors associated with bathtub drownings and may aid in the prevention of such deaths in the pediatric population.

Pediatric drowning: a 20-year review of autopsied cases: i. demographic features

Drowning is a significant cause of childhood mortality in the developed world. Males and preschool-aged children are overrepresented in most series of pediatric drowning. To obtain a better understanding of age- and gender-related factors, a retrospective review of autopsy records over a 20-year period (1984–2003) for cases of drowning (exclusive of bathtub drownings) was performed. Sixty-three consecutive cases of drowning were identified in 45 males and 18 females (ratio 2.5; P = 0.0007). The ages ranged from 9 months to 17 years, and most cases occurred in the 0- to 4-year age group (57%). Boys tended to be older than girls (5 years 9 months versus 4 years 4 months; P = 0.2133). Sites of drowning were open water (48%) and pools (44%), with children drowning in open water being significantly older than those drowning in pools (6 years 6 months versus 4 years 6 months; P = 0.0462). Most boys drowned in open water (53%), whereas most girls drowned in pools (50%). Resuscitation was attempted after the drowning episode in 84% of cases, with children aged 4 years or less having significantly higher rates than older children (P = 0.0346). The present study highlights differences in the pattern of drowning according to age and sex and may aid in developing targeted educational and preventative campaigns.

Pediatric drowning: a 20-year review of autopsied cases: II. Pathologic features

The pathologic findings in autopsies of drowning victims are nonspecific and vary from case to case. However, most reported pathologic series of drowning cases exclude children and do not take into consideration the unique circumstances surrounding bathtub drownings. In addition, the effect of resuscitation on the autopsy findings has not been studied in children. A retrospective review of autopsy records of nonbathtub drownings from a 20-year period (1984–2003) was performed and 63 cases were identified in 45 males and 18 females (age range 9 months to 17 years). The incidence of frothy exudate, pleural effusion, and increased lung weight was 43%, 36%, and 80%, respectively. The incidence of frothy exudate and the combination of all 3 factors was significantly higher in cases with no resuscitation compared with those cases with attempted resuscitation with or without delayed death. As the interval between the drowning episode and autopsy increased, the incidence of frothy exudate decreased significantly. There was no relationship between these findings and the age and sex of the decedent. Other clinical conditions or occult pathologic findings that may have contributed to death were found in 8 cases (13%). The findings highlight the need for thorough clinicopathologic correlation in cases of drowning to accurately interpret the pathologic findings.