Medical Query Research Articles

Improved Recognition of Maternal Deaths Using Modern Data Analytics.

To use a data-fusion approach to improve ascertainment of maternal deaths not detected with standard surveillance strategies. We conducted a retrospective cohort study from the electronic health records of a tertiary medical center from 2011 to 2018. Cases of maternal death were identified in two ways: 1) using a standard medical informatics service query of hospital data and 2) using the TriNetX discovery tool as patients with a vital status of "deceased" and evidence of antecedent pregnancy exposure based on such factors as obstetric diagnostic codes or obstetric-related procedures. Potential cases of maternal death identified by the latter method underwent chart review to confirm timing of death compared with timing of last appreciable pregnancy, and to characterize the details of these deaths. The primary outcome was pregnancy-associated mortality during pregnancy or within the first postpartum year in the discovery cohort compared with the hospital-identified cohort. Secondary outcomes included causes of death and comorbidities. During the study period, the standard service query identified 23 maternal deaths. The discovery tool identified 18 additional patients confirmed on subsequent chart review to represent pregnancy-associated deaths, a 78% increase in ascertainment of which a greater proportion represented postpartum deaths. The majority (61%) of newly ascertained mortalities were related to cardiac causes or other medical comorbidities. Although many hospital-ascertained cases were associated with deaths after delivery of a living newborn, more deaths after early pregnancy loss or termination were identified through the discovery tool. Improved recognition of pregnancy-associated deaths can be achieved with modern data analytics.

Probabilistic domain-knowledge modeling of disorder pathogenesis for dynamics forecasting of acute onset

Disease pathogenesis, a type of domain knowledge about biological mechanisms leading to diseases, has not been adequately encoded in machine-learning-based medical diagnostic models because of the inter-patient variabilities and complex dependencies of the underlying pathogenetic mechanisms. We propose 1) a novel pathogenesis probabilistic graphical model (PPGM) to quantify the dynamics underpinning patient-specific data and pathogenetic domain knowledge, 2) a Bayesian-based inference paradigm to answer the medical queries and forecast acute onsets. The PPGM model consists of two components: a Bayesian network of patient attributes and a temporal model of pathogenetic mechanisms. The model structure was reconstructed from expert knowledge elicitation, and its parameters were estimated using Variational Expectation-Maximization algorithms. We benchmarked our model with two well-established hidden Markov models (HMMs) – Input-output HMM (IO-HMM) and Switching Auto-Regressive HMM (SAR-HMM) – to evaluate the computational costs, forecasting performance, and execution time. Two case studies on Obstructive Sleep Apnea (OSA) and Paroxysmal Atrial Fibrillation (PAF) were used to validate the model. While the performance of the parameter learning step was equivalent to those of IO-HMM and SAR-HMM models, our model forecasting ability was outperforming those two models. The merits of the PPGM model are its representation capability to capture the dynamics of pathogenesis and perform medical inferences and its interpretability for physicians. The model has been used to perform medical queries and forecast the acute onset of OSA and PAF. Additional applications of the model include prognostic healthcare and preventive personalized treatments.

A hybrid medical text classification framework: Integrating attentive rule construction and neural network

The main objective of this work is to improve the quality and transparency of the medical text classification solutions. Conventional text classification methods provide users with only a restricted mechanism (based on frequency) for selecting features. In this paper, a three-stage hybrid method combining the gated attention-based bi-directional Long Short-Term Memory (ABLSTM) and the regular expression based classifier is proposed for medical text classification tasks. The bi-directional Long Short-Term Memory (LSTM) architecture with an attention layer allows the network to weigh words according to their perceived importance and focus on crucial parts of a sentence. Feature words (or keywords) extracted by ABLSTM model are utilized to guide the regular expression rule construction. Our proposed approach leverages the advantages of both the interpretability of rule-based algorithms and the computational power of deep learning approaches for a production-ready scenario. Experimental results on real-world medical online query data clearly validate the superiority of our system in selecting domain-specific and topic-related features. Results show that the proposed approach achieves an accuracy of 0.89 and an F1-score of 0.92 respectively. Furthermore, our experimentation also illustrates the versatility of regular expressions as a user-level tool for focusing on desired patterns and providing interpretable solutions for human modification.

Low Immunoglobulin Levels In Patients With Mastocytosis And Associated Clinical Manifestations

Patients with low IgG alone or in combination with low IgA or IgM have been reported to be susceptible to respiratory tract infections and recurrent sinusitis. Patients with CVID also have a higher prevalence of autoimmune diseases and lymphoid malignancies. Mastocytosis is a myeloproliferative disease that is not typically associated with autoimmune disease or frequent infections. We determined the prevalence of infectious and autoimmune disorders in patients with mastocytosis with low IgG alone or in combination with low IgM and/or IgA to ascertain if there was clinical relevance. We performed a 10 year retrospective analysis on 320 adult and pediatric patients with all variants of mastocytosis using an electronic medical query. We identified 26 adults and 9 children with one or more low immunoglobulins levels. These patient records were then examined for a history of infections and autoimmune disorders. Overall, among these 35 patients, 17% of patients had a history of infections and 11% had an autoimmune disorder. The most common infectious condition was recurrent otitis media (OM). No patients required hospitalization for infection, two patients met criteria for CVID but did not require replacement therapy. Only one pediatric patient with low IgG received IgG replacement therapy for OM and mastoiditis. The routine determination of immunoglobulin levels in mastocytosis is thus not required and should be reserved for those with a clinical condition which might relate to a low immunoglobulin levels.

Analysis of Factors Influencing Relapse and Pregnancy in Patients with Borderline Ovarian Tumors.

Objective: This retrospective study analyzed the factors affecting recurrence in patients after surgery with borderline ovarian tumors and postoperative recurrence and pregnancy after fertility-sparing surgery (FSS), to provide guidance for clinical treatment of borderline ovarian tumors and propose a therapeutic strategy for fertility protection.Methods: A total of 415 patients with borderline ovarian tumors were initially operated on in the gynecology ward of Shengjing Hospital Affiliated with China Medical University from September 1, 2013, to September 1, 2019. Central pathology review and prospective follow-up were carried out. The clinical and pathological data were consulted through the medical record query system of our hospital. The recurrence and pregnancy of the patients were investigated through telephone follow-up and outpatient and inpatient medical records. The influence of clinical and pathological variables on recurrence and pregnancy were evaluated using univariate/multivariate analyses.Results: In this study, 415 patients were collected, of which 21 lost follow-up, and a total of 394 eligible patients were included in the analysis. Among these patients, 25 patients relapsed with a recurrence rate of 6.3% and there were 196 patients with fertility-sparing surgery, of the 63 patients attempting to conceive, 35 were able to attain pregnancy with a pregnancy rate of 55.6%. All patients survived until the follow-up deadline. In univariate and multivariate analyses, FSS, FIGO stage, and micropapillary pattern were independent risk factors for recurrence of BOTs. FIGO stage, micropapillary pattern were independent risk factors for recurrence of BOTs with FSS. The risk of recurrence was not related to omentectomy nor postoperative chemotherapy. While omentectomy and chemotherapy had an impact on the pregnancy rate (P <0.05) and the pregnancy rate of patients without omentectomy or chemotherapy was higher.Conclusion: Omentectomy did not affect recurrence and it is not recommended as a routine operation. Adjuvant chemotherapy does not reduce the recurrence rate. While omentectomy and chemotherapy had an impact on the pregnancy rate, and both of them should be carried out more carefully in patients with fertility requirements.

Effect of penicillin allergy on prophylactic antibiotic administration in the parturient undergoing cesarean delivery.

Administering antibiotics is often difficult in patients with specific medication allergies. This investigation aimed to determine if a penicillin or cephalosporin allergy increased the risk for not receiving correct timing of prophylactic antibiotics at cesarean delivery. We hypothesized that patients with a penicillin or cephalosporin allergy would be less likely to receive antibiotics prior to incision for cesarean delivery. All women undergoing cesarean delivery at Mayo Clinic Hospital, from 1 March 2008, to 28 February 2018 were retrospectively identified by electronic medical record query. Patients were grouped based on allergy status to penicillin/cephalosporins. Data recorded included the type and time of antibiotic given in relation to surgical incision. The primary outcome of this study was administration of antibiotics within 60minutes prior to surgical incision. Characteristics potentially associated with the primary outcome were assessed using logistic regression. Of the 818 patients with a penicillin or cephalosporin allergy, 75 (9.2%) did not receive prophylactic antibiotic within 60minutes prior to skin incision. Conversely, 326 (6.9%) of the 4744 patients without a penicillin or cephalosporin allergy did not receive their prophylactic antibiotic within 60minutes prior to skin incision (P=.019). Patients undergoing an emergent cesarean delivery were also at an increased risk of not receiving their prophylactic antibiotic within 60minutes prior to skin incision (P<.001). Patients with a penicillin or cephalosporin allergy were less likely to receive prophylactic antibiotics within the recommended 60minutes prior to surgical incision. Clear plans and communication are important for ensuring proper antibiotic administration at cesarean delivery to prevent surgical site infection.

Abstract 16013: Evidence for Improvement in NYHA Class and Freedom From Hospitalization in Bariatric Surgery Patients With an Established Diagnosis of Heart Failure

Introduction: Heart failure patients with severe obesity endure significant morbidity and frequent hospitalizations. Bariatric surgery is proven to provide durable weight loss for those with severe obesity, but the clinical impact and safety of these procedures among patients with heart failure has not been well-demonstrated. Methods: We conducted a medical record query of patients who have a previous diagnosis of heart failure (HFpEF and HFrEF) and underwent subsequent Roux-En-Y gastric bypass or laparoscopic sleeve gastrectomy at a high-volume metabolic and bariatric surgery center. We compared clinical, demographic, and echocardiographic data captured just prior to the bariatric procedure to the most recent data available in the medical record for each patient. Results: There were 50 patients (88% had HFpEF) included in this study. Time from HF diagnosis to most recent follow-up ranged from 0.2 to 20.3 years (median 6.7 years) and there was no recorded mortality. The median time from HF diagnosis to surgery was 2.3 years and median time from surgery to recent follow-up was 2.9 years. Post-operative median decrease in BMI was 8.8 kg/m 2 , HF hospitalizations were 0.4 per patient year (PPY) to 0.15 PPY, p=0.008, and median NYHA Class was II pre-op and I post-op, p=0.048). LVEF, LVESD, and LVEDD were not significantly changed post-operatively (Table 1). Conclusion: Weight loss following bariatric surgery for patients with HF led to improvements in NYHA Class, fewer hospitalizations for HF, and was not associated with perioperative mortality. It is reasonable to consider bariatric surgery for this patient population, but further prospective investigation is warranted.

313 Outcomes for Patients With Congestive Heart Failure and Chronic Kidney Disease Receiving Fluid Resuscitation for Severe Sepsis or Septic Shock

Sepsis core measures are an integral part of sepsis treatment. Current fluid administration guidelines consist of administering at least 30cc/kg of intravenous fluids (IVF) per ideal body weight (IBW) within the first three hours of sepsis diagnosis regardless of pre-existing comorbidities at risk for fluid overload. This study aims to evaluate the outcomes of patients with a history of congestive heart failure (CHF) and/or chronic kidney disease (CKD) who receive fluid resuscitation for the management of severe sepsis or septic shock. We performed a retrospective case-control study of emergency department patients treated for severe sepsis or septic shock between April 2018 and May 2019. Subjects were identified from an electronic medical record query from 3 community hospitals, using ICD-10 codes. We identified subjects with a history of CHF or CKD (at-risk group [AR]) and a sample of patients without a history of CHF/ CKD (control group). We performed a structured chart review recording demographics, fluids received, airway interventions required, and outcome. Logistic regression analysis was used to compare the association between the amount of IV fluids received and the outcomes of interest (advance airway management and hospital mortality). Log-Rank test was used to adjust for multiple comparisons in survival analysis. Our cohort consisted of 745 patients with a history of CHF and/or CKD (AR group) and 570 patients without a history of CHF and/or CKD (control group). The mean age of patients in the AR group was 73.1 (SD 14.14) vs. 64.3 (16.9) in the control group (p-value <0.001). The initial mean lactate was similar in both the AR group= 3.02 (2.32) vs. the control group = 3.12 (2.44) (p-value=0.478). Overall patients in the AR group received less IVFs than the control group at 24 hours (2530.6 vs 3046.7, p-value=0.001). In the control group 50 (9%) vs 126 (17%) in the AR group required BiPAP during their hospitalization. There was a significant association between receipt of >30cc/kg of IVF in the AR group at 3hrs and 6hrs from ED arrival and the need for BiPAP (p-value=0.006, p=0.02, respectively). In-hospital mortality was found in 96 (13%) of the AR group vs 47 (8%) of the control group (p-value= 0.007). There was no statistically significant association between receipt of >30 cc/kg of IVF in the AR group compared to the control group at 3 or at 6 hrs from ED arrival in terms of in-hospital mortality (p-value=0.614, p=0.115, respectively). We identified a significant association of >30 cc/kg IVF administration and the need for Bi-PAP in AR patients. We identified higher in-hospital mortality in the AR group, but this was not associated with the amount of IVF resuscitation received. Further prospective studies that specifically identify at-risk populations receiving IVF resuscitation can better define the true risk for need for respiratory support.View Large Image Figure ViewerDownload Hi-res image Download (PPT)

A Parallel Particle Swarm Optimisation for Selecting Optimal Virtual Machine on Cloud Environment

Cloud computing has a significant role in healthcare services, especially in medical applications. In cloud computing, the best choice of virtual machines (Virtual_Ms) has an essential role in the quality improvement of cloud computing by minimising the execution time of medical queries from stakeholders and maximising utilisation of medicinal resources. Besides, the best choice of Virtual_Ms assists the stakeholders to reduce the total execution time of medical requests through turnaround time and maximise CPU utilisation and waiting time. For that, this paper introduces an optimisation model for medical applications using two distinct intelligent algorithms: genetic algorithm (GA) and parallel particle swarm optimisation (PPSO). In addition, a set of experiments was conducted to provide a competitive study between those two algorithms regarding the execution time, the data processing speed, and the system efficiency. The PPSO algorithm was implemented using the MATLAB tool. The results showed that the PPSO algorithm gives accurate outcomes better than the GA in terms of the execution time of medical queries and efficiency by 3.02% and 37.7%, respectively. Also, the PPSO algorithm has been implemented on the CloudSim package. The results displayed that the PPSO algorithm gives accurate outcomes better than default CloudSim in terms of final implementation time of medicinal queries by 33.3%. Finally, the proposed model outperformed the state-of-the-art methods in the literature review by a range from 13% to 67%.

Anesthesia and potential aerosol generation during magnetic resonance imaging in children with COVID-19.

The American College of Radiology recommends minimizing Magnetic Resonance Imaging (MRI) in COVID-19 patients, postponing non-urgent exams, and using alternative imaging.1 Sedation/anesthesia are aerosol generating procedures (AGP) due to requirement of bag-mask ventilation, intubation, and extubation with consequent risk of exposure to healthcare workers. This is complicated by limitation in the use of personal protective equipment (PPE) in the magnet zone (Zone IV).We describe our experience for children requiring anesthesia for emergency MRI during the COVID-19 outbreak in Philadelphia.

Survey on Information Retrieval Accuracy in Biomedical Systems

In recent years, research articles in the biomedicine domain have increased rapidly which makes it difficult for medical personnel to manually gather all the information required. To meet this information, need better, Information Retrieval (IR) techniques have been employed for accurate and timely search of the information. A special lexicon set identifies the medical domain and hence requires specialized processing steps to handle complex medical terminologies. The query and result mapping may need extra attention in order to give best results. The queries may be in the form of short questions, medical case reports, medical case narratives, verbose medical queries, community questioning, semi-structured queries, etc. The paper intends to explore the work done in the field of medical text information retrieval. Models developed or employed, indexing and ranking is surveyed and comparison is done among them.

Adverse Drug Events due to Central Nervous System Depressant Drugs in Pediatric Patients With or Without Surgery.

To identify differences in the incidence and severity of adverse drug events (ADEs) due to CNS depressant drugs among pediatric patients with and without surgery. The Japan Adverse Drug Events Study was a cohort study enrolling pediatric inpatients. Potential ADEs were identified by onsite review of medical charts, incident reports, and prescription queries. Two independent physicians classified ADEs and severity. We compared the incidence and characteristics of ADEs between pediatric patients with surgery (surgery group) and without surgery (non-surgery group). We evaluated severity of ADEs due to CNS depressant drugs among both groups. We enrolled 944 patients, 234 in surgery group and 710 in non-surgery group. A total of 480 ADEs due to any drugs occurred in 225 patients. Among 81 ADEs due to CNS depressant drugs, 42 ADEs were in surgery group, whereas 39 were in non-surgery group. The risk of fatal or life-threatening ADEs due to CNS depressant drugs was significantly higher than other drugs (12% vs. 2%, p < 0.001). In the surgery group, anesthetics led to 2 fatal or life-threatening, 8 serious, and 30 significant ADEs, whereas in the non-surgery group anesthetics led to 2 fatal or life-threatening, 5 serious, and 4 significant ADEs. Anesthetics were higher risk in the non-surgery group (p = 0.049). The risks of fatal and life-threatening ADEs were significantly higher with CNS depressant drugs than other drugs. Pediatric patients without surgery have higher risks of fatal or life-threatening ADEs due to anesthetics than those with surgery.

Understanding Medical Conversations with Scattered Keyword Attention and Weak Supervision from Responses

In this work, we consider the medical slot filling problem, i.e., the problem of converting medical queries into structured representations which is a challenging task. We analyze the effectiveness of two points: scattered keywords in user utterances and weak supervision with responses. We approach the medical slot filling as a multi-label classification problem with label-embedding attentive model to pay more attention to scattered medical keywords and learn the classification models by weak-supervision from responses. To evaluate the approaches, we annotate a medical slot filling data and collect a large scale unlabeled data. The experiments demonstrate that these two points are promising to improve the task.

Opioid stewardship: implementing a proactive, pharmacist-led intervention for patients coprescribed opioids and benzodiazepines at an urban academic primary care centre

In 2017, almost 4000 Canadians died from opioid-related causes. Coadministration of opioids and benzodiazepines is a risk factor for overdose. Few studies have evaluated leveraging pharmacists to address opioid-benzodiazepine coprescribing....

Neuroendocrine metastases to the ovaries are significantly associated with small bowel neuroendocrine tumors and carcinomatosis

IntroductionNeuroendocrine tumors (NETs) metastatic to the ovary are traditionally considered rare, but data are lacking. This study seeks to better characterize the prevalence and outcomes of patients with neuroendocrine ovarian metastases (NOM). MethodsWomen with well-differentiated lung and gastroenteropancreatic NETs 2007–2017 were identified by medical record query. Clinicopathologic data were reviewed among patients with and without NOM. ResultsOf 242 patients, 27 (11.2%) developed NOM. NOM developed in 24.8% of SBNET patients and 65.7% of patients with carcinomatosis and intact ovaries. 33.3% had associated small bowel obstructions; 11.1% had ureteral obstruction. NOM were not apparent on imaging in 29.6% nor visible intraoperatively in 8.3%. Five-year survival rate was 61.5%. Those who underwent oophorectomy had a lower rate of subsequent ureteral obstruction (p < 0.01). ConclusionsNOM are more prevalent than previously reported and associated with significant morbidity. Empiric oophorectomy may be considered for SBNET patients and strongly advised in carcinomatosis.

Google Medical Update: Why Is the Search Engine Decreasing Visibility of Health and Medical Information Websites?

The Google search engine answers many health and medical information queries every day. People have become used to searching for this type of information. This paper presents a study which examined the visibility of health and medical information websites. The purpose of this study was to find out why Google is decreasing the visibility of such websites and how to measure this decrease. Since August 2018, Google has been more rigorously rating these websites, since they can potentially impact people’s health. The method of the study was to collect data about the visibility of health and medical information websites in sequential time snapshots. Visibility consists of combined data of unique keywords, positions, and URL results. The sample under study was made up of 21 websites selected from 10 European countries. The findings reveal that in sequential time snapshots, search visibility decreased. The decrease was not dependent on the country or the language. The main reason why Google is decreasing the visibility of such websites is that they do not meet high ranking criteria.

Methylprednisolone-related liver injury: A descriptive study using the French pharmacovigilance database

Hepatotoxicity associated with methylprednisolone (MP) is rarely reported in the literature. The aim of the present study was to review the characteristics of acute liver injury associated with intravenous (IV) or oral MP registered in the French pharmacovigilance database (FPD). All cases with MP coded as suspected, concomitant, or interacting drug associated with liver injury as the adverse effect reported up to May 2016 were extracted from the FPD. Cases were identified using the "Drug related hepatic disorders" Standard Medical Query. A total of 97 cases of liver injury associated with MP were analysed; 58.8% were women and the median age was 46 years (range: 1-91). MP was used for an autoimmune disease in 47.6% of cases including 26 cases of multiple sclerosis, and was IV in 79.4% of cases. Nearly three-quarters of patients (73,2%) had a hepatocellular type of injury, the severity of which was mainly mild (45%) or moderate (31%). Most patients (92%) spontaneously and fully recovered within a mean 38.4 days. A rechallenge using the IV route was performed in 13 patients and for 10 (76.9%) this was positive (the initial type of injury was hepatocellular for all these cases). Regarding IV route of administration (n=77), MP was coded as the only suspected drug in 22% of cases. The results suggest that IV MP causality should be considered in case of acute liver injury while data for oral MP is insufficient; systematic liver monitoring for high-dose IV MP may be recommended.

Head Concepts Selection for Verbose Medical Queries Expansion

Semantic concepts and relations encoded in domain-specific ontologies and other medical semantic resources play a crucial role in deciphering terms in medical queries and documents. The exploitation of these resources for tackling the semantic gap issue has been widely studied in the literature. However, there are challenges that hinder their widespread use in real-world applications. Among these challenges is the insufficient knowledge individually encoded in existing medical ontologies, which is magnified when users express their information needs using long-winded natural language queries. In this context, many of the users query terms are either unrecognized by the used ontologies, or cause retrieving false positives that degrade the quality of current medical information search approaches. In this article, we explore the combination of multiple extrinsic semantic resources in the development of a full-fledged medical information search framework to: i) highlight and expand head medical concepts in verbose medical queries (i.e. concepts among query terms that significantly contribute to the informativeness and intent of a given query), ii) build semantically enhanced inverted index documents, iii) contribute to a heuristical weighting technique in the query document matching process. To demonstrate the effectiveness of the proposed approach, we conducted several experiments over the CLEF eHealth 2014 dataset. Findings indicate that the proposed method combining several extrinsic semantic resources proved to be more effective than related approaches in terms of precision measure.

Primary Care Access to an Online Decision Support Tool is Associated With Improvements in Some Aspects of Pediatric Migraine Care.

To evaluate whether primary care provider (PCP) access to an online decision support tool is associated with a change in evidence-based primary care medical management of pediatric migraine. In this prospective observational study, PCPs serving a target community were educated on the availability and use of an online clinical decision support tool that was developed to inform treatment of pediatric migraine. For 9 months before and after implementation of the decision tool, the proportions of children with migraine prescribed evidence-based and contraindicated medications by PCPs in the target region were monitored using electronic medical record query and statistically compared to these same proportions for patients in surrounding (control) regions. Rates of visits to the emergency department for migraine also were tracked pre- and postimplementation as an indirect measure of impact of the decision tool. Provider usage of the decision tool was monitored and summarized using web analytics. Approximately half (56%) of target region PCPs used the online tool at least once over the project period. Relative to control regions and baseline trends, the proportion of children residing in the target region who were prescribed recommended abortive and preventive medications for treating migraine was statistically significantly higher following implementation of the tool. No significant changes to frequency of emergency care visits for migraine by youth in the target region were observed. Availability to PCPs of an online decision support tool for pediatric migraine is associated with a modest change in some aspects of evidence-based medical care.

Analyzing How Elderly Acute Myeloid Leukemia Patients Are Treated, an Institutional Experience

Rationale: Treatment of acute myelogenous leukemia (AML) remains a challenge in elderly populations, those with comorbid conditions, and patients with poor performance status indices. The optimal choice for induction therapy as well as further agent selection is unclear, and current guidelines recommend enrollment in a controlled clinical trial. Methods: Institutional cases of AML via an electronic medical record query performed in November 2017 containing cases of AML in patients 65 years of age or greater from 01/01/2000 to 06/21/2017 were extracted. Instances of acute myelogenous leukemia were identified by ICD codes. Age, gender, induction therapy, cytogenetics, molecular analyses, and overall survival were collected. Results: A total of 61 cases of AML in patients aged 65 or greater were identified, with those having incomplete data being excluded from analysis. The mean age of included patients was 78.9 years of age, 35 were male and 26 were female. 60 confirmed deaths were recorded. 26 patients received conventional 7+3 (42.6%), 22 received a hypomethylating agent (HMA) (31.1%), 16 (26.2%) did not receive treatment. Conclusion: Our institutional data showed overall survival was significantly longer when treated with 7+3, 354 days (95% CI [93, 614]), vs. 61 days (95% CI [15, 107]). Similarly, OS was significantly longer when treated with HMA, 303 days (95% CI [23, 583]). Risk of death, accordingly, was as follows: 7+3 HR .347 (.179-.672), HMA HR .348 (.173-.703). Our institutional mortality data reasonably reflected SEER data analysis reported by Medeiros et al concerning untreated patients, but trended toward a greater OS for those treated with 7+3 or HMA. Registry data, as well as our institutional data, demonstrate a survival benefit to intensive chemotherapy and palliative chemotherapy, while controlled trials have not shown this benefit consistently in elderly populations. Practices regarding induction therapy vary greatly between institutions. Determination of which elderly patients to treat with intensive therapy remains difficult.