The promises and challenges of clinical AI in community paediatric medicine.

The widespread adoption of virtual care technologies has quickly reshaped healthcare operations and delivery, particularly in the context of community medicine. In this paper, we use the virtual care landscape as a point of departure to envision the promises and challenges of artificial intelligence (AI) in healthcare. Our analysis is directed towards community care practitioners interested in learning more about how AI can change their practice along with the critical considerations required to integrate AI into their practice. We highlight examples of how AI can enable access to new sources of clinical data while augmenting clinical workflows and healthcare delivery. AI can help optimize how and when care is delivered by community practitioners while also improving practice efficiency, accessibility, and the overall quality of care. Unlike virtual care, however, AI is still missing many of the key enablers required to facilitate adoption into the community care landscape and there are challenges we must consider and resolve for AI to successfully improve healthcare delivery. We discuss several critical considerations, including data governance in the clinic setting, healthcare practitioner education, regulation of AI in healthcare, clinician reimbursement, and access to both technology and the internet.

Imported pediatric malaria at the Hospital for Sick Children, Toronto, Canada: a 16 year review.

BACKGROUND: Children under 5 represent 86% of annual malaria deaths in the world. Following increasing trends in international travel, cases of imported malaria are rising in North America. We describe the epidemiology of malaria diagnosed at a tertiary care pediatric center in the multicultural city of Toronto. METHOD: Retrospective chart review of all laboratory confirmed malaria from birth to <18 years between July 1, 1997 and June 30, 2013. Epidemiological data, travel history, chemoprophylaxis history, as well as clinical presentation, diagnosis and treatment were extracted. RESULTS: In total 107 children were diagnosed with malaria in the 16 year time period. Plasmodium falciparum malaria was identified in 76 (71%), Plasmodium vivax in 28 (26%). Median age of infected children was 6.7 years where 35% of children were born in Canada, 63% were recent or previous immigrants. Of those who resided in Canada, reason for travel included visiting friends or relatives (VFR) 95% and tourism or education (5%). Most common countries of infection were Ghana (22%), Nigeria (20%) and India (14%). Median parasitemia at presentation to our institution was 0.4% (IQR 0.1-2.3) with a maximum parasitemia of 31%. Nineteen (18%) met the WHO criteria for severe malaria due to hyperparasitemia, with 3 of these cases also meeting clinical criteria for severe malaria. One third of patients had a delay in treatment of 2 or more days. Ten percent of children had seen two or more primary health care professionals prior to admission. Prophylaxis was documented in 22 (21%), and out of those, 6 (27%) were appropriate for the region of travel and only 1 case was documented as adherent to their prescription. There were no cases of fatality. CONCLUSION: Malaria continues to be a significant disease in returning travelers and immigrant or refugee populations. An increase in physician awareness is required. Appropriate pre-travel advice, insect protection measures, effective chemoprophylaxis is needed to reduce the incidence and improve the management of imported pediatric malaria.

Does this child have bacterial meningitis? A systematic review of clinical prediction rules for children with suspected bacterial meningitis.

BACKGROUND: Acute meningitis is a relatively common phenomenon in children. Identifying which children are most likely to have bacterial meningitis vs. self-limiting aseptic meningitis is important, as these children require investigation and antibiotic treatment. OBJECTIVE: Our aim was to systematically identify and review the quality and performance of published clinical prediction rules (CPRs) for children with suspected bacterial meningitis. METHODS: Medline and Embase were searched for CPRs involving children 0-18 years of age with suspected bacterial meningitis, with cerebral spinal fluid (CSF) culture used as the reference diagnostic standard. CPR quality was assessed using 17 previously published items. CPR performance was evaluated using sensitivity, negative likelihood ratio, and the treatment frequency that would result if the rule was used. RESULTS: Eleven studies involving 6675 children with acute meningitis fulfilled all inclusion criteria and were entered in the study. They all describe the derivation or validation of six unique CPRs. A rigorously developed, high-performing, and well-validated CPR ready for clinical use to guide which children with suspected bacterial meningitis should be hospitalized and treated with intravenous antibiotics and which can be safely discharged home was not identified. Areas for quality improvement for future CPR studies include prospective validation using standardized inclusion criteria, adequate blinding, predictor reproducibility assessment, and meticulous follow-up of outcomes. The Bacterial Meningitis Score had the highest quality and performance and is the best candidate for prospective validation. CONCLUSIONS: Until consistently high methodological quality and diagnostic performance are demonstrated through prospective validation, caution is warranted in the routine clinical use of existing CPRs for children with suspected bacterial meningitis.

Pulse probiotic administration induces repeated small intestinal Muc3 expression in rats.

Upon ingestion, probiotics may act to protect the host through a number of protective mechanisms including modulation of genes involved in intestinal innate mucosal defense such as epithelial cell-derived mucin glycoproteins and inhibitor of apoptosis proteins. To determine the specificity of effect and sustainability of response in vivo, Lactobacillus plantarum 299v (Lp299v), Lactobacillus rhamnosus R0011 (LrR0011), and Bifidobacterium bifidum R0071 (BbR0071) were added repeatedly or intermittently to the drinking water of Sprague-Dawley rats. After killing the rats via CO2 suffocation, Muc2, Muc3, neuronal apoptosis inhibitor protein (NAIP), human inhibitor of apoptosis protein 1/cellular inhibitor of apoptosis 2 (HIAP1/cIAP2), and human inhibitor of apoptosis protein 2/cellular inhibitor of apoptosis 1 (HIAP2/cIAP1) mRNA and protein levels were analyzed via RT-PCR and immunohistochemistry. Live Lp299v, BbR0071, and LrR0011 increased Muc3 protein and mRNA expression in jejunum and ileum. Heat-killed and a nonadherent derivative of Lp299v failed to induce Muc3 expression. Lp299v did induce expression of HIAP2/cIAP1 and NAIP expression. Muc3 mucin expression was elevated for 5 d after oral administration of Lp299v; however, this effect was not sustained despite ongoing daily ingestion of a probiotic. Intermittent pulse ingestion of probiotics, however, was found to repeatedly increase Muc3 expression. We conclude that selected probiotics can induce protective genes of mucosal intestinal epithelial cells, an effect that is reproducible with pulse probiotic administration.

Predictors of Invasive Herpes Simplex Virus Infection in Young Infants.

OBJECTIVES: To identify independent predictors of and derive a risk score for invasive herpes simplex virus (HSV) infection. METHODS: In this 23-center nested case-control study, we matched 149 infants with HSV to 1340 controls; all were ≤60 days old and had cerebrospinal fluid obtained within 24 hours of presentation or had HSV detected. The primary and secondary outcomes were invasive (disseminated or central nervous system) or any HSV infection, respectively. RESULTS: Of all infants included, 90 (60.4%) had invasive and 59 (39.6%) had skin, eyes, and mouth disease. Predictors independently associated with invasive HSV included younger age (adjusted odds ratio [aOR]: 9.1 [95% confidence interval (CI): 3.4-24.5] <14 and 6.4 [95% CI: 2.3 to 17.8] 14-28 days, respectively, compared with >28 days), prematurity (aOR: 2.3, 95% CI: 1.1 to 5.1), seizure at home (aOR: 6.1, 95% CI: 2.3 to 16.4), ill appearance (aOR: 4.2, 95% CI: 2.0 to 8.4), abnormal triage temperature (aOR: 2.9, 95% CI: 1.6 to 5.3), vesicular rash (aOR: 54.8, (95% CI: 16.6 to 180.9), thrombocytopenia (aOR: 4.4, 95% CI: 1.6 to 12.4), and cerebrospinal fluid pleocytosis (aOR: 3.5, 95% CI: 1.2 to 10.0). These variables were transformed to derive the HSV risk score (point range 0-17). Infants with invasive HSV had a higher median score (6, interquartile range: 4-8) than those without invasive HSV (3, interquartile range: 1.5-4), with an area under the curve for invasive HSV disease of 0.85 (95% CI: 0.80-0.91). When using a cut-point of ≥3, the HSV risk score had a sensitivity of 95.6% (95% CI: 84.9% to 99.5%), specificity of 40.1% (95% CI: 36.8% to 43.6%), and positive likelihood ratio 1.60 (95% CI: 1.5 to 1.7) and negative likelihood ratio 0.11 (95% CI: 0.03 to 0.43). CONCLUSIONS: A novel HSV risk score identified infants at extremely low risk for invasive HSV who may not require routine testing or empirical treatment.

Performance of the Modified Boston and Philadelphia Criteria for Invasive Bacterial Infections.

BACKGROUND: The ability of the decades-old Boston and Philadelphia criteria to accurately identify infants at low risk for serious bacterial infections has not been recently reevaluated. METHODS: We assembled a multicenter cohort of infants 29 to 60 days of age who had cerebrospinal fluid (CSF) and blood cultures obtained. We report the performance of the modified Boston criteria (peripheral white blood cell count [WBC] ≥20 000 cells per mm3, CSF WBC ≥10 cells per mm3, and urinalysis with >10 WBC per high-power field or positive urine dip result) and modified Philadelphia criteria (peripheral WBC ≥15 000 cells per mm3, CSF WBC ≥8 cells per mm3, positive CSF Gram-stain result, and urinalysis with >10 WBC per high-power field or positive urine dip result) for the identification of invasive bacterial infections (IBIs). We defined IBI as bacterial meningitis (growth of pathogenic bacteria from CSF culture) or bacteremia (growth from blood culture). RESULTS: We applied the modified Boston criteria to 8344 infants and the modified Philadelphia criteria to 8131 infants. The modified Boston criteria identified 133 of the 212 infants with IBI (sensitivity 62.7% [95% confidence interval (CI) 55.9% to 69.3%] and specificity 59.2% [95% CI 58.1% to 60.2%]), and the modified Philadelphia criteria identified 157 of the 219 infants with IBI (sensitivity 71.7% [95% CI 65.2% to 77.6%] and specificity 46.1% [95% CI 45.0% to 47.2%]). The modified Boston and Philadelphia criteria misclassified 17 of 53 (32.1%) and 13 of 56 (23.3%) infants with bacterial meningitis, respectively. CONCLUSIONS: The modified Boston and Philadelphia criteria misclassified a substantial number of infants 29 to 60 days old with IBI, including those with bacterial meningitis.

Cerebrospinal Fluid Reference Values for Young Infants Undergoing Lumbar Puncture.

OBJECTIVES: To determine age-specific reference values and quantify age-related changes for cerebrospinal fluid (CSF) white blood cell (WBC) counts and protein and glucose concentrations in infants ≤60 days of age. METHODS: This multicenter, cross-sectional study included infants ≤60 days old with CSF cultures and complete CSF profiles obtained within 24 hours of presentation. Those with conditions suspected or known to cause abnormal CSF parameters (eg, meningitis) and those with a hospital length of stay of >72 hours were excluded. Reference standards were determined for infants ≤28 days of age and 29 to 60 days of age by using the third quartile +1.5 interquartile range for WBC and protein and the first quartile -1.5 interquartile range for glucose. CSF parameter centile curves based on age were calculated by using the LMST method. RESULTS: A total of 7766 patients were included. CSF WBC counts were higher in infants ≤28 days of age (upper bound: 15 cells/mm3) than in infants 29 to 60 days of age (upper bound: 9 cells/mm3; P < .001). CSF protein concentrations were higher in infants ≤28 days of age (upper bound: 127 mg/dL) than in infants 29 to 60 days of age (upper bound: 99 mg/dL; P < .001). CSF glucose concentrations were lower in infants ≤28 days of age (lower bound: 25 mg/dL) than in infants 29 to 60 days of age (lower bound: 27 mg/dL; P < .001). CONCLUSIONS: The age-specific CSF WBC count, protein concentration, and glucose concentration reference values identified in this large, multicenter cohort of infants can be used to interpret the results of lumbar puncture in infants ≤60 days of age.

Herpes Simplex Virus Infection in Infants Undergoing Meningitis Evaluation.

BACKGROUND: Although neonatal herpes simplex virus (HSV) is a potentially devastating infection requiring prompt evaluation and treatment, large-scale assessments of the frequency in potentially infected infants have not been performed. METHODS: We performed a retrospective cross-sectional study of infants ≤60 days old who had cerebrospinal fluid culture testing performed in 1 of 23 participating North American emergency departments. HSV infection was defined by a positive HSV polymerase chain reaction or viral culture. The primary outcome was the proportion of encounters in which HSV infection was identified. Secondary outcomes included frequency of central nervous system (CNS) and disseminated HSV, and HSV testing and treatment patterns. RESULTS: Of 26 533 eligible encounters, 112 infants had HSV identified (0.42%, 95% confidence interval [CI]: 0.35%-0.51%). Of these, 90 (80.4%) occurred in weeks 1 to 4, 10 (8.9%) in weeks 5 to 6, and 12 (10.7%) in weeks 7 to 9. The median age of HSV-infected infants was 14 days (interquartile range: 9-24 days). HSV infection was more common in 0 to 28-day-old infants compared with 29- to 60-day-old infants (odds ratio 3.9; 95% CI: 2.4-6.2). Sixty-eight (0.26%, 95% CI: 0.21%-0.33%) had CNS or disseminated HSV. The proportion of infants tested for HSV (35%; range 14%-72%) and to whom acyclovir was administered (23%; range 4%-53%) varied widely across sites. CONCLUSIONS: An HSV infection was uncommon in young infants evaluated for CNS infection, particularly in the second month of life. Evidence-based approaches to the evaluation for HSV in young infants are needed.

Correction of Cerebrospinal Fluid Protein in Infants With Traumatic Lumbar Punctures.

In our multicenter cohort of infants ≤60 days of age, we identified 2646 infants with a traumatic lumbar puncture, of which 31 (1.2%) had bacterial meningitis. For every 1000 cerebrospinal fluid red blood cells/mm, cerebrospinal (cerebrospinal fluid) protein increased 1.1 mg/dL (95% confidence interval: 1.0-1.2 mg/dL).

Concomitant Bacterial Meningitis in Infants With Urinary Tract Infection.

To determine age-stratified prevalence of concomitant bacterial meningitis in infants ≤60 days with a urinary tract infection, we performed a 23-center, retrospective study of 1737 infants with urinary tract infection. Concomitant bacterial meningitis was rare, but more common in infants 0-28 days of age [0.9%; 95% confidence interval (CI): 0.4%-1.9%) compared with infants 29-60 days of age (0.2%; 95% CI: 0%-0.8%).

Properties of serial ultrasound clinical diagnostic pathway in suspected appendicitis and related computed tomography use.

OBJECTIVES: The primary objective was to determine the diagnostic accuracy of a serial ultrasound (US) clinical diagnostic pathway to detect appendicitis in children presenting to the emergency department (ED). The secondary objective was to examine the diagnostic performance of the initial and interval US and to compare the accuracy of the pathway to that of the initial US. METHODS: This was a prospective cohort study of 294 previously healthy children 4 to 17 years old with suspected appendicitis and baseline pediatric appendicitis scores of ≥2, who were managed with the serial US clinical diagnostic pathway. This pathway consisted of an initial US followed by a clinical reassessment in each patient and an interval US and surgical consultation in patients with equivocal initial US and persistent concern about appendicitis. The USs were interpreted by published criteria as positive, negative, or equivocal for appendicitis. Children in whom this pathway did not rule in or rule out appendicitis underwent computed tomography (CT). Cases with missed appendicitis, negative operations, and CTs after the pathway were considered inaccurate. The primary outcome was the diagnostic accuracy of the serial US clinical diagnostic pathway. The secondary outcomes included the test performance of the initial and interval US imaging studies. RESULTS: Of the 294 study children, 111 (38%) had appendicitis. Using the serial US clinical diagnostic pathway, 274 of 294 children (93%, 95% confidence interval [CI] = 90% to 96%) had diagnostically accurate results: 108 of the 111 (97%) appendicitis cases were successfully identified by the pathway without CT scans (two missed and one CT), and 166 of the 183 (91%) negative cases were ruled out without CT scans (14 negative operations and three CTs). The sensitivity of this pathway was 108 of 111 (97%, 95% CI = 94% to 100%), specificity 166 of 183 (91%, 95% CI = 87% to 95%), positive predictive value 108 of 125 (86%; 95% CI = 79% to 92%), and negative predictive value 166 of 169 (98%, 95% CI = 96% to 100%). The diagnostic accuracy of the pathway was higher than that of the initial US alone (274 of 294 vs. 160 of 294; p < 0.0001). Of 123 patients with equivocal initial US, concern about appendicitis subsided on clinical reassessment in 73 (no surgery and no missed appendicitis). Of 50 children with persistent symptoms, 40 underwent interval US and 10 had surgical consultation alone. The interval US confirmed or ruled out appendicitis in 22 of 40 children (55.0%) with equivocal initial US, with one false-positive interval US. CONCLUSIONS: The serial US clinical diagnostic pathway in suspected appendicitis has an acceptable diagnostic accuracy that is significantly higher than that of the initial US and results in few CT scans. This approach appears most useful in children with equivocal initial US, in whom the majority of negative cases were identified at clinical reassessment and appendicitis was diagnosed by interval US or surgical consultation in most study patients.

Does this child have appendicitis? A systematic review of clinical prediction rules for children with acute abdominal pain.

OBJECTIVES: To systematically identify clinical prediction rules (CPRs) for children with suspected appendicitis and compare their methodological quality and performance. STUDY DESIGN AND SETTING: Included studies involved children aged 0-18 years with suspected appendicitis identified through MEDLINE and EMBASE from 1950 to 2012. The quality was assessed using 17 previously published items. The performance was evaluated using the sensitivity, negative likelihood ratio, and predicted frequency of appendicitis diagnosis that would result if the rule was used. RESULTS: Twelve studies fulfilled the inclusion criteria describing the derivation or validation of six unique CPRs involving 4,201 children with suspected appendicitis. Migratory pain, nausea or vomiting, and right lower quadrant tenderness were common predictors to all rules. Methodological quality varied widely. The most poorly addressed quality items were the predictor and outcome assessor blinding, predictor description, and reproducibility of predictor assessment. The most well-validated CPRs were the Pediatric Appendicitis Score (PAS) and MANTRELS (Migration, Anorexia, Nausea/vomiting, Tenderness in the right lower quadrant, Rebound pain, Elevation in temperature, Leukocytosis, Shift to the left)/Alvarado Score. Overall, the PAS validation studies outperformed the Alvarado validation studies. CONCLUSION: The PAS and Alvarado scores were the most well validated but neither met the current performance benchmarks. A high quality, well validated, and consistently high-performing CPR was not identified. Further research is needed before a CPR for children with suspected appendicitis can be used in routine practice.

Should a hospitalized child receive empiric treatment with acyclovir?

BACKGROUND: Herpes simplex encephalitis is associated with substantial morbidity and mortality and may be related to timely diagnosis and treatment. While awaiting the results of testing, hospitalization and empiric treatment with acyclovir is recommended, though the direct and indirect costs associated with this management are substantial. We sought to examine children hospitalized for possible herpes simplex encephalitis, following clinical and laboratory assessment in the emergency department, and empiric treatment with acyclovir, in order to describe the proportion receiving a complete course of treatment; and to identify the clinical variables which are associated with receiving a complete course, as compared with an incomplete course of acyclovir. METHODS: Hospitalized children prescribed acyclovir were included in this case control study. Clinical, laboratory and diagnostic variables were abstracted for children prescribed a complete (≥ 14 days) or an incomplete course (<14 days) of acyclovir. Odds ratios and 95% confidence intervals were calculated. RESULTS: 289 children met eligibility criteria, 30 (10%) received a complete course and 259 (90%) received an incomplete course. A history of mucocutaneous herpes simplex virus infection (p < 0.01), Glasgow Coma Scale ≤ 13 (p = 0.02), focal neurologic findings (p = 0.001) and elevated cerebrospinal fluid white blood cell count (p = 0.05) were associated with a complete course of acyclovir. CONCLUSIONS: Many children did not complete a full course of therapy. Unnecessary testing and treatment is burdensome to families and the health care system. Possible predictive variables include abnormal Glascow Coma Scale, focal neurologic findings and cerebrospinal fluid pleocytosis.

Homeless youth's overwhelming health burden: A review of the literature.

Homelessness has reached epidemic proportions in Canada. Canadian children and adolescents are the most vulnerable because youth comprise the fastest growing segment of the homeless population. A systematic literature review was undertaken using MEDLINE, Web of Science and the Homeless Hub (www.homelesshub.ca) to encompass the time frame from January 1990 to June 2009. The following terms were used as key words: 'homelessness', 'homeless youth', 'poverty', 'street youth' and 'runaway'. The present review identified an intersection among education deficits, social service insufficiencies, and poor mental and physical health in homeless youth. Health care delivery to homeless youth was often nonanticipatory, inconsistent and perceived as discriminatory. However, street youth were identified as requiring health care for pregnancy, mental health concerns, sexually transmitted illnesses, respiratory conditions, substance abuse and a myriad of other illnesses. Plenty of work is still required to reduce health inequalities and improve the daily living conditions of Canadian youth living in poverty.

Clinical prediction rules for children: a systematic review.

CONTEXT: The degree to which clinical prediction rules (CPRs) for children meet published standards is unclear. OBJECTIVE: To systematically review the quality, performance, and validation of published CPRs for children, compare them with adult CPRs, and suggest pediatric-specific changes to CPR methodology. METHODS: Medline was searched from 1950 to 2011. Studies were selected if they included the development of a CPR involving children younger than 18 years. Two investigators assessed study quality, rule performance, and rule validation as methodologic standards. RESULTS: Of 7298 titles and abstracts assessed, 137 eligible studies were identified. They describe the development of 101 CPRs addressing 36 pediatric conditions. Quality standards met in fewer than half of the studies were blind assessment of predictors (47%), reproducibility of predictors (18%), blind assessment of outcomes (42%), adequate follow-up of outcomes (36%), adequate power (43%), adequate reporting of results (49%), and 95% confidence intervals reported (36%). For rule performance, 48% had a sensitivity greater than 0.95, and 43% had a negative likelihood ratio less than 0.1. For rule validation, 76% had no validation, 17% had narrow validation, 8% had broad validation, and none had impact analysis performed. Compared with CPRs for adult health conditions, quality and rule validation seem to be lower. CONCLUSIONS: Many CPRs have been derived for children, but few have been validated. Relative to adult CPRs, several quality indicators demonstrated weaknesses. Existing performance standards may prove elusive for CPRs that involve children. CPRs for children that are more assistive and less directive and include patients' values and preferences in decision-making may be helpful.

Prospective evaluation of residents on call: before and after duty-hour reduction.

BACKGROUND: On July 1, 2009, in Ontario the maximum period of continuous duty that residents were permitted to work was reduced from 28 to 24 hours. We evaluated the effect of regulation on residents in 3 eras: 2 before (2005 and early 2009) and 1 after (late 2009) the duty-hour reduction. METHODS: On-call pediatric residents on pediatric medicine rotations prospectively recorded the numbers of patients (assigned and admitted) and the durations of direct patient care, documentation, staff supervision, and education attended. Sleep was measured with actigraphy. RESULTS: The 51 residents worked 180 duty periods, were assigned a median of 6 (interquartile range: 4 -12) daytime patients and 24 (interquartile range: 19-30) overnight patients. Residents reported spending means of 239 minutes providing direct patient care, 235 minutes documenting, and 243 minutes sleeping and receiving 73 minutes of staff supervision and 52 minutes of education. From early 2009 to after duty-hour reduction, residents provided 47 fewer (19.6%) minutes of direct patient care (P = .056) and received 44 fewer minutes (60.3%) of supervision (P = .0005) but spent similar times documenting, receiving education, and sleeping. In early 2009, residents provided 73 more minutes (30.5%) of direct patient care (P = .0016), spent 63 more minutes (26.8%) documenting, and slept 105 fewer minutes (43.0%) (P = .0062) than in 2005. DISCUSSION: After duty-hour reduction in 2009, we found reduced supervision and direct patient care. Comparison of the 2 periods before duty-hour reduction showed less sleep and longer patient contact in early 2009, which suggests that changes occurred without regulation.