Husham Sharifi is a Clinical Assistant Professor in Stanford's Division of Pulmonary, Allergy, and Critical Care Medicine. He sees patients in the Intensive Care Unit and in a specialty clinic for individuals with pulmonary manifestations after stem cell transplant, as well as patients with persistent symptoms after infection with COVID-19 (i.e., long COVID). His research interests include economics, policy, and data science in medicine. He holds a masters degree in Technology and Policy from MIT, a masters degree in Bioinformatics from Stanford, and an MD from Yale.
- Pulmonary Disease
Clinical Assistant Professor, Medicine - Pulmonary, Allergy & Critical Care Medicine
Fellowship, Stanford University Medical Center, Pulmonary and Critical Care
Residency, Yale New Haven Hospital, Internal Medicine
MD, Yale University School of Medicine
MS, Stanford University, Bioinformatics
MS, Massachusetts Institute of Technology, Engineering
MS, Massachusetts Institute of Technology, Technology and Policy
Head-to-head Comparison of Qualitative Radiologist Assessment With Automated Quantitative Computed Tomography Analysis for Bronchiolitis Obliterans Syndrome After Hematopoietic Cell Transplantation.
Journal of thoracic imaging
PURPOSE: Computed tomography (CT) findings of bronchiolitis obliterans syndrome (BOS) can be nonspecific and variable. This study aims to measure the incremental value of automated quantitative lung CT analysis to clinical CT interpretation. A head-to-head comparison of quantitative CT lung density analysis by parametric response mapping (PRM) with qualitative radiologist performance in BOS diagnosis was performed.MATERIALS AND METHODS: Inspiratory and end-expiratory CTs of 65 patients referred to a post-bone marrow transplant lung graft-versus-host-disease clinic were reviewed by 3 thoracic radiologists for the presence of mosaic attenuation, centrilobular opacities, airways dilation, and bronchial wall thickening. Radiologists' majority consensus diagnosis of BOS was compared with automated PRM air trapping quantification and to the gold-standard diagnosis of BOS as per National Institutes of Health (NIH) consensus criteria.RESULTS: Using a previously established threshold of 28% air trapping on PRM, the diagnostic performance for BOS was as follows: sensitivity 56% and specificity 94% (area under the receiver operator curve [AUC]=0.75). Radiologist review of inspiratory CT images alone resulted in a sensitivity of 80% and a specificity of 69% (AUC=0.74). When radiologists assessed both inspiratory and end-expiratory CT images in combination, the sensitivity was 92% and the specificity was 59% (AUC=0.75). The highest performance was observed when the quantitative PRM report was reviewed alongside inspiratory and end-expiratory CT images, with a sensitivity of 92% and a specificity of 73% (AUC=0.83).CONCLUSIONS: In the CT diagnosis of BOS, qualitative expert radiologist interpretation was noninferior to quantitative PRM. The highest level of diagnostic performance was achieved by the combination of quantitative PRM measurements with qualitative image feature assessments.
View details for DOI 10.1097/RTI.0000000000000595
View details for PubMedID 33999570
The Use of Factor Eight Inhibitor Bypass Activity (FEIBA) for the Treatment of Perioperative Hemorrhage in Left Ventricular Assist Device Implantation.
Journal of cardiothoracic and vascular anesthesia
OBJECTIVE: To test the hypothesis that factor eight inhibitor bypassing activity (FEIBA) can be used to control bleeding following left ventricular assist device (LVAD) implantation without increasing the 14-day composite thrombotic outcome of pump thrombus, ischemic cerebrovascular accidents, pulmonary embolism, and deep venous thrombosis.DESIGN: Retrospective cohort study.SETTING: Academic hospital.PARTICIPANTS: Three hundred nineteen consecutive patients who underwent LVAD implantation (December 1, 2009 to December 30, 2018).INTERVENTION: FEIBA administered to control perioperative hemorrhage.MEASUREMENTS AND MAIN RESULTS: The 82 patients (25.7%) in the FEIBA cohort had more risk factors for perioperative hemorrhage, such as lower preoperative platelet count (169 ± 66 v 194 ± 68 * 103/mL, p = 0.004), prior cardiac surgery (36.6% v 21.9%, p = 0.008), and longer cardiopulmonary bypass (CPB) time (100.3 v 75.2 minutes, p = 0.001) than the 237 controls. After 16.6 units (95% CI: 14.3-18.9) of blood products were given, 992 units (95% CI: 821-1163) of FEIBA were required to control bleeding in the FEIBA cohort. Compared to the controls, there were no differences in the 14-day composite thrombotic outcome (11.0% v 7.6%, p = 0.343) or mortality rate (3.7% v 1.3%, p = 0.179). Multivariate logistical regression identified preoperative international normalized ratio (odds ratio [OR]: 1.30, 95% CI: 1.04-1.62) and CPB time (OR: 1.11, 95% CI: 1.02-1.20) as risk factors for 14-day thrombotic events, but FEIBA usage was not associated with an increased risk.CONCLUSIONS: In this retrospective cohort study, the use of FEIBA (1,000 units, 13 units/kg) to control perioperative hemorrhage following LVAD implantation was not associated with increases in mortality or composite thrombotic outcome.
View details for DOI 10.1053/j.jvca.2021.04.030
View details for PubMedID 34034934
- Radiology-pathology Correlation in Recovered COVID-19, Demonstrating Organizing Pneumonia. American journal of respiratory and critical care medicine 2020
Machine learning algorithms to differentiate among pulmonary complications after hematopoietic cell transplant.
Pulmonary complications, including infections, are highly prevalent in patients after hematopoietic cell transplant with chronic graft-versus-host disease. These comorbid diseases can make the diagnosis of early lung graft-versus-host disease (bronchiolitis obliterans syndrome) challenging. A quantitative method to differentiate among these pulmonary diseases can address diagnostic challenges and facilitate earlier and more targeted therapy.We conducted a single center study of 66 patients with computed tomography chest scans analyzed with a quantitative imaging tool known as parametric response mapping. Parametric response mapping results were correlated with pulmonary function tests and clinical characteristics. Five parametric response mapping metrics were applied to K-means clustering and support vector machine models to distinguish among post-transplant lung complications solely from quantitative output.Compared to parametric response mapping, spirometry showed a moderate correlation with radiographic air trapping, and total lung capacity and residual volume showed a strong correlation with radiographic lung volumes. K-means clustering analysis distinguished 4 unique clusters. Clusters 2 and 3 represented obstructive physiology (encompassing 81% of patients with bronchiolitis obliterans syndrome) in increasing severity (percent air trapping 15.6% and 43.0%, respectively). Cluster 1 was dominated by normal lung, and cluster 4 was characterized by patients with parenchymal opacities. A support vector machine algorithm differentiated bronchiolitis obliterans syndrome with specificity of 88%, sensitivity of 83%, accuracy of 86% and an area under the receiver operating characteristic curve of 0.85.Our machine learning models offer a quantitative approach for the identification of bronchiolitis obliterans syndrome versus other lung diseases, including late pulmonary complications after hematopoietic cell transplant.
View details for DOI 10.1016/j.chest.2020.02.076
View details for PubMedID 32343962
Predicting the Incidence of Pressure Ulcers in the Intensive Care Unit Using Machine Learning.
EGEMS (Washington, DC)
2019; 7 (1): 49
Background: Reducing hospital-acquired pressure ulcers (PUs) in intensive care units (ICUs) has emerged as an important quality metric for health systems internationally. Limited work has been done to characterize the profile of PUs in the ICU using observational data from the electronic health record (EHR). Consequently, there are limited EHR-based prognostic tools for determining a patient's risk of PU development, with most institutions relying on nurse-calculated risk scores such as the Braden score to identify high-risk patients.Methods and Results: Using EHR data from 50,851 admissions in a tertiary ICU (MIMIC-III), we show that the prevalence of PUs at stage 2 or above is 7.8 percent. For the 1,690 admissions where a PU was recorded on day 2 or beyond, we evaluated the prognostic value of the Braden score measured within the first 24 hours. A high-risk Braden score (<=12) had precision 0.09 and recall 0.50 for the future development of a PU. We trained a range of machine learning algorithms using demographic parameters, diagnosis codes, laboratory values and vitals available from the EHR within the first 24 hours. A weighted linear regression model showed precision 0.09 and recall 0.71 for future PU development. Classifier performance was not improved by integrating Braden score elements into the model.Conclusion: We demonstrate that an EHR-based model can outperform the Braden score as a screening tool for PUs. This may be a useful tool for automatic risk stratification early in an admission, helping to guide quality protocols in the ICU, including the allocation and timing of prophylactic interventions.
View details for DOI 10.5334/egems.307
View details for PubMedID 31534981
Association between Inhaled Corticosteroid Use and Pulmonary Nontuberculous Mycobacterial Infection
ANNALS OF THE AMERICAN THORACIC SOCIETY
2018; 15 (10): 1169–76
Nontuberculous mycobacterial (NTM) pulmonary disease prevalence is increasing.To determine the association between the use of inhaled corticosteroids and the likelihood of NTM pulmonary infection among individuals with treated airway disease.We conducted a case-control study of subjects with airway disease with and without NTM pulmonary infection (based on mycobacterial respiratory cultures) between 2000 and 2010 in northern California. We quantified the use of inhaled corticosteroids, other airway disease medications, and healthcare use within 6 months of NTM pulmonary infection identification. We used 1:10 case-control matching and conditional logistic regression to evaluate the association between the duration and cumulative dosage of inhaled corticosteroid use and NTM pulmonary infection.We identified 248 cases with NTM pulmonary infection with an estimated rate of 16.4 cases per 10,000 subjects treated for airway disease. The median interval between treated airway disease cohort entry (defined as date of patient filling the third airway disease treatment prescription) and NTM case identification was 1,217 days. Compared with control subjects, subjects with NTM pulmonary infection were more likely to use airway disease medications including systemic steroids; they were also more likely to use health care. Any inhaled corticosteroids use between 120 days and 2 years before cohort entry was associated with substantially increased odds of NTM infection. For example, the adjusted odds ratio for NTM infection among inhaled corticosteroid users in a 2-year interval was 2.51 (95% confidence interval, 1.40-4.49; P < 0.01). Increasing cumulative inhaled corticosteroid dose was also associated with greater odds of NTM infection.Inhaled corticosteroid use, and particularly high-dose inhaled corticosteroid use, was associated with an increased risk of NTM pulmonary infection.
View details for PubMedID 30213194
Analysis of Hypoxic and Hypercapnic Ventilatory Response in Healthy Volunteers
2017; 12 (1)
A previous study has suggested that the Human Leukocyte Antigen (HLA) allele DQB1*06:02 affects hypoxic ventilatory response (HVR) but not hypercapnic ventilatory response (HCVR) in an Asian population. The current study evaluated the relationship in Caucasians and Asians. In addition we assessed whether gender or polymorphisms in genes participating in the control of breathing affect HVR and HCVR.A re-breathing system was used to measure HVR and HCVR in 551 young adults (56.8% Caucasians, 30% Asians). HLA-DQB1*06:02 and tagged polymorphisms and coding variants in genes participating in breathing (PHOX2B, GPR4 and TASK2/KCNK5) were analyzed. The associations between HVR/HCVR and HLA-DQB1*06:02, genetic polymorphisms, and gender were evaluated using ANOVA or frequentist association testing with SNPTEST.HVR and gender are strongly correlated. HCVR and gender are not. Mean HVR in women was 0.276±0.168 (liter/minute/%SpO2) compared to 0.429±0.266 (liter/minute/%SpO2) in men, p<0.001 (55.4% higher HVR in men). Women had lower baseline minute ventilation (8.08±2.36 l/m vs. 10.00±3.43l/m, p<0.001), higher SpO2 (98.0±1.3% vs. 96.6±1.7%, p<0.001), and lower EtCO2 (4.65±0.68% vs. 4.82±1.02%, p = 0.025). One hundred and two (18.5%) of the participants had HLA-DQB1*06:02. No association was seen between HLA-DQB1*06:02 and HVR or HCVR. Genetic analysis revealed point wise, uncorrected significant associations between two TASK2/KCNK5 variants (rs2815118 and rs150380866) and HCVR.This is the largest study to date reporting the relationship between gender and HVR/ HCVR and the first study assessing the association between genetic polymorphisms in humans and HVR/HCVR. The data suggest that gender has a large effect on hypoxic breathing response.
View details for DOI 10.1371/journal.pone.0168930
View details for Web of Science ID 000391612300073
View details for PubMedID 28045995
View details for PubMedCentralID PMC5207520
Cost-effectiveness analysis of periacetabular osteotomy.
journal of bone and joint surgery. American volume
2008; 90 (7): 1447-1456
A lack of long-term outcomes data following periacetabular osteotomy makes it difficult for surgeons to recommend the most appropriate procedure to young patients who might be candidates for a joint-preserving procedure. In this study, we compared the cost-effectiveness of periacetabular osteotomy with total hip arthroplasty in terms of cost per quality-adjusted life year for the young adult.A decision model was constructed for a cost-utility analysis of periacetabular osteotomy compared with total hip arthroplasty. Outcome probabilities and effectiveness were derived from the literature. Effectiveness was expressed in quality-adjusted life years gained. Cost data were compiled and verified from our institution. Costs and utilities were discounted in accord with the United States Panel on Cost-Effectiveness in Health and Medicine. Principal outcome measures were average incremental costs, incremental effectiveness, incremental quality-adjusted life years, and net health benefits. Multivariate sensitivity analysis was used to assess the contribution of included variables in the model's outcomes.For Tönnis grade-1 coxarthrosis, periacetabular osteotomy dominates with an average incremental cost-effectiveness of $7856 per quality-adjusted life year and an average incremental effectiveness of 0.15. For Tönnis grade-2 coxarthrosis, periacetabular osteotomy is, on the average, more cost-effective than total hip arthroplasty with an incremental cost-effectiveness of $824 per quality-adjusted life year, but it is less effective than total hip arthroplasty, on the average, with an incremental effectiveness of -1.4 quality-adjusted life years. Periacetabular osteotomy becomes more cost-effective at a longevity of 5.5 years for Tönnis grade-1 coxarthrosis and 18.25 years for Tönnis grade-2 coxarthrosis. In Tönnis grade-3 coxarthrosis, total hip replacement becomes the dominant treatment strategy.Periacetabular osteotomy is, on the average, more cost-effective in Tönnis grade-1 and grade-2 coxarthrosis, while it is both more costly and less effective in Tönnis grade-3 coxarthrosis. These findings can inform clinical decision-making in the absence of long-term data. On the basis of this model, periacetabular osteotomy is preferable to total hip arthroplasty in Tönnis grade-1 and grade-2 coxarthrosis when the patient is sufficiently young and when functionality in sports is important.
View details for DOI 10.2106/JBJS.G.00730
View details for PubMedID 18594092
Cost-effectiveness analysis of unicompartmental knee arthroplasty as an alternative to total knee arthroplasty for unicompartmental osteoarthritis.
journal of bone and joint surgery. American volume
2006; 88 (9): 1975-1982
The purpose of this study was to compare the cost-effectiveness of unicompartmental knee arthroplasty as an alternative to total knee arthroplasty in patients with degenerative arthritis limited to either the medial or lateral compartment.A decision model was created for the treatment of end-stage unicompartmental knee arthritis. A literature review was used to identify possible outcomes and their probabilities following treatment with either unicompartmental or total knee arthroplasty. Each outcome was weighted for quality of life with use of a utility factor, and effectiveness was expressed in units of quality-adjusted life years. Gross costs were estimated from Medicare reimbursement data for the relevant Current Procedural Terminology and Diagnosis-Related Group codes.Sensitivity analysis demonstrated that the cost-effectiveness of unicompartmental knee arthroplasty is dependent on the assumption that its durability and functional outcomes approach those of total knee arthroplasty. Specifically, it is necessary for the survival of unicompartmental implants to be within three to four years of the assumed survival of total knee implants for unicompartmental arthroplasty to remain a cost-effective alternative. Under these assumptions, the use of unicompartmental arthroplasty is a cost-effective choice as it results in incremental gains in effectiveness at a cost of less than US dollars 50,000 (in 1998 United States dollars) per quality-adjusted life year gained.This study supports unicompartmental knee arthroplasty as a cost-effective alternative for the treatment of unicompartmental arthritis when the durability and function of a unicompartmental replacement are assumed to be similar to those of a primary total knee replacement. This suggests that, with appropriate patient selection, the currently available literature supports unicompartmental arthroplasty as a cost-effective alternative to total knee arthroplasty.Economic and decision analysis, Level II.
View details for PubMedID 16951114
Cost-effectiveness analysis of core decompression
JOURNAL OF ARTHROPLASTY
2006; 21 (5): 670-681
Core decompression is widely used to treat the early stages of osteonecrosis of the hip. The purpose of this analysis is to assist orthopedic surgeons in judging whether currently available data support the use of core decompression as cost-effective. A decision model was created for the treatment of osteonecrosis of the femoral head. Literature review was used to identify possible outcomes and their probability after initial treatment with either observation or core decompression. This model demonstrates core decompression must delay the need for total hip arthroplasty for a minimum of 5 years to maintain an incremental cost-effectiveness ratio lower than 50,000 dollars per quality-adjusted life year gained. Treatment options with ratios higher than 50,000 dollars per quality-adjusted life year are generally considered to have limited cost-effectiveness. This study demonstrates that core decompression has the potential to be a highly cost-effective alternative if it is leads to a delay in the need for total hip arthroplasty of 5 years or longer.
View details for DOI 10.1016/j.arth.2005.08.018
View details for Web of Science ID 000239835800008
View details for PubMedID 16877152