Lehana Thabane’s research while affiliated with St. Joseph's Healthcare Hamilton and other places

Background Community-acquired pneumonia (CAP) in young children is mainly caused by viral pathogens; for this reason, Infectious Disease Society of America guidelines explicitly state that antibiotic treatment of non-severe CAP in this population is not obligatory. However, the vast majority of non-severe pediatric CAP episodes diagnosed in North America prompt antibiotic prescription anyway. Methods A before-after prospective study was done in a Canadian emergency department (ED) enrolling children aged 6+ months without comorbidities diagnosed with CAP and managed as outpatients. In the before (control) phase, participants were managed as per routine care. In the after (intervention) phase, participants were managed using a novel care pathway. This care pathway used imaging results, a point-of-care C-reactive protein (POC CRP) assay, and other already-available data (eg. O2 saturation, respiratory rate, age) to categorize participants as ‘appreciable risk’ or ‘low risk.’ Those at ‘appreciable risk’ were given a prescription for antibiotics, whereas those at ‘low risk’ were simply discharged home. The next day, caregivers of ‘low-risk’ participants were informed of results of nasopharyngeal viral/atypical pathogen testing. Follow-up was done at day 2-5, 14-21, and 30. Results There were 84 participants recruited in the control phase (Mar-Nov 2022) and 78 in the intervention phase (Nov 2022-Mar 2024). Control phase participants had a median age of 2.8 y (IQR 2.1 – 4.4y) and 13% had normal chest imaging, whereas intervention phase participants had a median age of 3.7 y (IQR 1.9 – 6 y) and 5.1% had normal chest imaging. All 84 of the participants in the control phase were prescribed antibiotics at ED discharge, as compared to only 51% of those in the intervention phase (49% less, 95%CI 38-60% less, p< 0.001). There were no significant differences in clinical cure at day 2-5, re-presentation to the ED, later hospitalization, or caregiver satisfaction with the care plan. Conclusion A simple novel care pathway incorporating POC CRP testing as well as other available data facilitated a very clinically important decrease in antibiotic prescribing without compromising clinical outcomes. This care pathway should be evaluated in the context of a multicentre randomized trial. Disclosures Jeffrey Pernica, MD, MSc, FRCPC, DTMH, MedImmune: Grant/Research Support|Merck: Grant/Research Support Dominik Mertz, MD, MSc, KCI Inc. USA: Grant/Research Support

Background Classic teaching is that spinal anesthesia is safe at or below the L2–L3 interspace. To evaluate this, we sought to determine the percentage of individuals with a conus medullaris termination (CMT) level at or below the L1–L2 interspace. Further, the relationship of CMT level to age, sex, body mass index (BMI), and spinal pathology was examined, as was the reliability of using Tuffier's line (TL) as an anatomical landmark. Methods This retrospective study evaluated magnetic resonance images of 944 adult patients to determine the CMT level. The relationship between age, sex, height, BMI, and spinal pathology and CMT level was explored by logistic regression. The correspondence of the TL line to the L4–L5 interspace and the presence of overlap with the CMT were examined using 720 lumbar x‐rays of the same patient cohort. Results Of 944 patients (mean age, 57.8 years; 49% male), 18.9% had CMT at or below the L1–L2 interspace, and spinal anesthesia at the L2–L3 interspace was found to carry a 0.7% incidence of neuraxial risk. Only the presence of congenital spinal abnormalities was found to be significantly predictive of having a CMT at or below the L1–L2 interspace. TL was found to correspond to the L4–L5 interspace in 99.8% of patients with lumbar x‐rays. Conclusions Spinal anesthesia at the L2–L3 interspace, using TL as an anatomical landmark, is safe in >99% of patients. However, caution must be exercised in all patients as demographic variables were found to be limited in predicting a low CMT level. Editorial Comment Unlike previous smaller studies, this retrospective study included MRI data from a total of 944 patients. The present study confirms that spinal anesthesia at the L2–L3 interspace or below can be considered safe. The findings indicate that Tuffier's line can be used as a reliable anatomical landmark.

Background As per IDSA guidelines, throat swab testing for group A Streptococcus is important to ensure appropriate antimicrobial use for children with pharyngitis that is not obviously viral. Bacterial culture is the diagnostic reference standard but takes days to result. With culture, physicians often initially provide a prescription that is only to be filled later on if GAS testing is positive. Unfortunately, this strategy can result in inappropriate treatment if parents of children with negative tests fill the script anyway, or if parents of children with positive tests cannot be contacted later. Point-of-care (POC) antigen testing in children is specific but not sensitive, so a negative test still requires culture verification, which limits its overall usefulness. POC molecular GAS testing is as sensitive as culture, and provides a result in minutes, which simplifies prescribing. ED physician satisfaction at enrolment Satisfaction was captured using a 5-point Likert scale on the day of enrolment. Methods A randomized controlled trial was done in a Canadian children’s emergency department (ED). Children aged >3 y with throat swabs ordered to diagnose GAS pharyngitis were eligible. Participants were randomized 1:1 to either the IDNOW Strep A2 (intervention) or bacterial culture (control). The primary outcome was appropriate treatment, ie taking antibiotics if GAS-positive or not taking antibiotics if GAS-negative. Secondary outcomes included physician and parent satisfaction and time to symptom resolution. Caregiver satisfaction at enrolment Satisfaction was captured using a 5-point Likert scale on the day of enrolment. Results 392 participants were recruited from Jan 2021 – Mar 2024. Intervention and control groups were similar in terms of mean age (8.2 vs 8.3 y) and mean MacIsaac score (both 3.3); 29% of the intervention group and 41% of the control group were GAS-positive. Participants in the intervention group were significantly more likely to receive appropriate antibiotics (90.8% vs 76.3%, OR 3.1 [95%CI 1.7-5.7], p< 0.001). ED physicians (pOR 147, p< 0.001) and parents (pOR 2.4, p< 0.001) were also much more likely to be satisfied with POC molecular testing. There were no statistically significant differences in time to resolution of symptoms, though comparisons favoured the intervention. Conclusion In the largest RCT of molecular GAS testing yet done, the intervention was associated with substantial improvements in appropriate antibiotic prescribing and was clearly preferred by both ED physicians and parents. Disclosures Jeffrey Pernica, MD, MSc, FRCPC, DTMH, MedImmune: Grant/Research Support|Merck: Grant/Research Support

Background Despite efforts to promote optimal breastfeeding practices, the practice of exclusive breastfeeding is low in South Africa. We conducted a trial to determine whether text messaging plus motivational interviewing prolonged exclusive breastfeeding during the first six months of life and improved child health outcomes. Methods We conducted a randomized parallel group-controlled trial between July 2022 and May 2024, at a secondary-level healthcare facility. Mothers living with HIV, 18 years or older, initiating breastfeeding, on combination antiretroviral therapy (cART) and their infants were enrolled. The primary endpoint was exclusive breastfeeding from birth through week 24, based on the consecutive 24-hour food recall interviews. We compared differences in exclusive breastfeeding rates using a proportion test. Mothers who completely stopped breastfeeding were asked an open-ended question on reasons for stopping breastfeeding. Results Using block randomization mother-child pairs (n = 276) were randomly allocated to receive intervention (n = 138) or standard infant feeding counselling (n = 138), of whom 105 and 101 mother-child pairs in the intervention group and standard care group, respectively, completed all four study visits. Exclusive breastfeeding rate at 24 weeks in the intervention group was 6% (6/105) and 7% (7/101) in the standard care group, rate difference − 1% (95% CI -6–4%). Sixty-two of 276 mothers completely stopped breastfeeding, of whom 25% (34/138) and 20% (28) were in the intervention group and standard care group, respectively. The most common reasons for stopping breastfeeding were the mother needing to return to work or look for work, 66% (n = 41). We also found that early breastfeeding cessation increased risk of child hospitalization or death compared to any form of breastfeeding to week 24, 10% (5/48) versus 3% (5/158), p = 0.055. Conclusions We found no effect of the intervention on exclusive breastfeeding rates. Early cessation of breastfeeding was prevalent and maternal employment characteristics are important social determinants of breastfeeding behaviour. There is need for further research evaluating the effect of interventions that include financial incentives on breastfeeding practices among socioeconomically disadvantaged mothers. HIV services should reliably offer cART, consistently monitor viral load, and support mothers cART adherence, in settings where mixed feeding is common. Trial registration The trial was registered on ClinicalTrials.gov (NCT05063240) and Pan African Clinical Trial Registries (PACTR202110870407786) before recruitment of the first subject.

Background Coronavirus disease 2019 (COVID-19) caused by severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) continues to have a significant impact worldwide, in part due to a reduction in neutralising antibody protection provided by vaccines targeting monovalent spike antigens related to immune escape. Development of vaccines amenable for respiratory mucosal delivery that provide broad and durable immunity are needed. Objective This study aims to determine the safety and immunogenicity of a single inhaled dose of a replication-deficient chimpanzee adenovirus type 68 vector expressing a trivalent transgene cassette of SARS-CoV-2 S1 domain of spike, nucleocapsid, and RNA polymerase genes (ChAd-triCoV/Mac). Methods We plan to recruit 350 non-pregnant adults aged 18–65 who have previously received three intramuscular SARS-CoV-2 mRNA vaccines in this 24-week, multicentre, double-blind, parallel group, phase II, two-sided superiority randomised controlled trial. Participants will be randomised 2:1 to receive either a single inhaled dose of ChAd-triCoV/Mac or placebo, both aerosolized via the AeroNeb® Solo vibrating mesh nebuliser. A subset of separately-randomised participants will undergo bronchoalveolar lavage (BAL). The co-primary outcome to be analysed in the per-protocol population is SARS-CoV-2 antigen-specific CD4+ and CD8+ T cell responses measured at 2 weeks in the peripheral blood; solicited adverse events (AEs) frequency to day 7 and unsolicited to day 28. In the BAL sub-study, the co-primary outcome will also include BAL fluid SARS-CoV-2 antigen-specific CD4+ and CD8+ T cell responses measured at 4 weeks. Conclusion This placebo-controlled phase 2 randomised trial will test the safety and immunogenicity of an inhaled, nebulised ChAd-triCoV/Mac SARS-CoV-2 vaccine that is novel in its administration route and targeting of multiple viral epitopes. The results will provide further information regarding the mucosal T cell response to immunisation.

Introduction Ultrarare diseases (URD) represent a challenge to health technology assessment (HTA). The traditional framework for assessing efficacy and cost effectiveness may be biased to include clinically relevant outcomes, leaving patient-centered outcomes doomed to neglect. Here we explore patient-centered outcomes in the context of patient and citizen involvement in the assessment of URD by the Brazilian National Committee for Health Technology Incorporation (CONITEC). Methods We assessed 53 HTA reports from CONITEC that evaluated URD-related technologies (and included highlights of patients’ and citizens’ perspectives during recommendation meetings) published from 2012 to 2022. Data extraction was performed by two independent researchers. Data on year of report, sex, ethnicity, category (patient or family), and previous experience with the assessed technology were extracted and analyzed using descriptive statistics. Patients’ and citizens’ narratives were collated from the reports. A thematic analysis was conducted according to patient-centered outcomes and technology-related outcomes and was then compared with the evidence synthesis protocol described in the HTA. Results Only seven URD-related HTA reports registered patient or citizen participation, all of which were published in 2022. The age of two participants was reported (both 17 years). Six participants were women. Ethnicity was not reported. All participants had previous experience with the technology. Four participants were family or caregivers and three were patients. Considering patient-centered outcomes, physical (muscular strength) and emotional (self-confidence) improvements that positively affected independence in basic daily functions were reported. These functions included activities such as dressing, self-care, cooking, and leisure. Advantages listed for the assessed technologies included the possibility of self-administration of medication (e.g., swallowing a pill, opening a medicine bottle, and using a syringe). Conclusions The results show that although, in some cases, primary outcomes reported in evidence synthesis protocols include patient-centered outcomes (e.g., activities of daily living), in other cases the evidence synthesis failed to identify relevant studies. In other cases, the reports failed to differentiate between primary and secondary outcomes or to fully account for patient-centered outcomes.

Introduction The decision-making process for incorporating technologies for ultrarare diseases (URD) has been a challenge for health technology assessment agencies worldwide. These challenges have been presented in debates about the budget impact of incorporating technologies for URD. This is an important issue because there are other dimensions of the economic and social impact of URD that require consideration. Methods Data were extracted from National Committee for Health Technology Incorporation (CONITEC) reports (2012 to 2022) on technologies for the treatment of URD in Brazil. Diseases were classified using an epidemiological criterion or Orphanet consultation (prevalence ≤1 per 50,000 inhabitants). Variables included eligible patient count, population estimation method, incremental impact values for one and five years, and diffusion rate in the first and fifth year. Univariate logistic regression was used to adjust the relationship between the budget impact analysis and the final recommendation, considering factors associated with incorporation in univariate regression and p-values less than 0.10 in a multivariate regression. Results Among 53 reports, 48 percent exclusively employed the epidemiological approach for incremental impact assessment population estimation, rising to 69.5 percent when combined with measured demand. Population data were nearly evenly sourced from national and international platforms, with the UK, the USA, and multicenter studies being the most cited internationally. Notable differences were found between favorable and unfavorable CONITEC recommendations, with lower values being associated with incorporation. Market share diffusion rates favored the option of 100 percent diffusion in both the first year and the cumulative five years. The analysis highlighted the influence of demand characteristics and technology type on the budget impact value over one and five years. Conclusions The study found that budget impact data significantly influenced the final recommendation for technology incorporation, indicating a criterion favoring technologies with a lower budget impact. However, requester characteristics and technology type also played a role in the decision-making process, suggesting that additional factors influence recommendations.

Introduction Several countries established health technology assessment (HTA) processes to support decision-making. Considering the high volume of submissions processed by HTA agencies, approaches to determine factors associated with the approval would be beneficial. This study aimed to predict the final recommendation of the National Committee for Health Technology Incorporation (Conitec) using a natural language processing (NLP) algorithm for text extraction. Methods Conitec’s 2012 to 2022 reports (n=389) were split into 75 percent training and 25 percent testing data. Tokenization enabled NLP models: Least Absolute Shrinkage and Selection Operator (LASSO), logistic regression, support vector machine (SVM), random forest, neural network, and Extreme Gradient Boosting (XGBOOST). Evaluation criteria included accuracy, area under the receiver operating characteristic curve (ROC AUC) score, precision, and recall. Cluster analysis with k-modes identified two clusters (group 0 = approved, group 1 = rejected). Results The neural network model demonstrated the best accuracy metrics with a precision of 0.815, accuracy of 0.769, ROC AUC of 0.871, and a recall of 0.746. Some tokenization identified that linguistic markers could contribute to the prediction of incorporation decision by the Brazilian HTA Committee, such as international HTA agencies’ experience and the government as the main requester. Cluster and XGBOOST analysis identified similar results with approved technologies with a predominance of drugs assessment, mainly requested by the government, and not approved mostly assessing drugs, the industry as the main requester. Conclusions The NLP model could identify predictors for the final decision process on the incorporation of health technologies in Brazil’s Unified Health System, opening paths for future work using HTA reports coming from other agencies. This model could potentially improve the throughput of HTA systems by supporting experts with prediction/factors/criteria for approval or nonapproval as an earlier step.