ICD-11 extension codes support detailed clinical abstraction and comprehensive classification

Background The new International Classification of Diseases—11th revision (ICD-11) succeeds ICD-10. In the three decades since ICD-10 was released, demands for detailed information on the clinical history of a morbid patient have increased. Methods ICD-11 has now implemented an addendum chapter X called “Extension Codes”. This chapter contains numerous codes containing information on concepts including disease stage, severity, histopathology, medicaments, and anatomical details. When linked to a stem code representing a clinical state, the extension codes add significant detail and allow for multidimensional coding. Results This paper discusses the purposes and uses of extension codes and presents three examples of how extension codes can be used in coding clinical detail. Conclusion ICD-11 with its extension codes implemented has the potential to improve precision and evidence based health care worldwide.

Distance-based Classification and Regression Trees for the analysis of complex predictors in health and medical research

Predicting patient outcomes based on patient characteristics and care processes is a common task in medical research. Such predictive features are often multifaceted and complex, and are usually simplified into one or more scalar variables to facilitate statistical analysis. This process, while necessary, results in a loss of important clinical detail. While this loss may be prevented by using distance-based predictive methods which better represent complex healthcare features, the statistical literature on such methods is limited, and the range of tools facilitating distance-based analysis is substantially smaller than those of other methods. Consequently, medical researchers must choose to either reduce complex predictive features to scalar variables to facilitate analysis, or instead use a limited number of distance-based predictive methods which may not fulfil the needs of the analysis problem at hand. We address this limitation by developing a Distance-Based extension of Classification and Regression Trees (DB-CART) capable of making distance-based predictions of categorical, ordinal and numeric patient outcomes. We also demonstrate how this extension is compatible with other extensions to CART, including a recently published method for predicting care trajectories in chronic disease. We demonstrate DB-CART by using it to expand upon previously published dose–response analysis of stroke rehabilitation data. Our method identified additional detail not captured by the previously published analysis, reinforcing previous conclusions. We also demonstrate how by combining DB-CART with other extensions to CART, the method is capable of making predictions about complex, multifaceted outcome data based on complex, multifaceted predictive features.

Validation of acute myocardial infarction (AMI) in electronic medical records: the SPEED-EXTRACT Study

ObjectivesTo determine whether data captured in electronic medical records (eMR) is sufficient to serve as a clinical data source to make a reliable determination of ST elevation myocardial infarction (STEMI) and non-ST elevation myocardial infarction (NSTEMI) and to use these eMR derived diagnoses to validate ICD-10 codes for STEMI and NSTEMI.DesignRetrospective validation by blind chart review of a purposive sample of patients with a troponin test result, ECG record, and medical note available in the eMR.SettingTwo local health districts containing two tertiary hospitals and six referral hospitals in New South Wales, Australia.ParticipantsN = 897 adult patients who had a hs-troponin test result indicating suspected AMI.Primary outcome measuresInter-rater reliability of clinical diagnosis (κ) for ST-elevated myocardial infarction (STEMI) and Non-ST elevated myocardial infarction (NSTEMI); and sensitivity, specificity, and positive predictive value (PPV) of ICD-10 codes for STEMI and NSTEMI.ResultsThe diagnostic agreement between clinical experts was high for STEMI (κ = 0.786) but lower for NSTEMI (κ = 0.548). ICD-10 STEMI codes had moderate sensitivity (Se = 88±6.7), very high specificity (Sp = 99±0.7) and high positive predictive value (PPV = 91±6). NSTEMI ICD-10 codes were lower in each case (Se = 69±6.4, Sp = 96.0±1.5, PPV = 84±6).ConclusionsThe eMR held sufficient clinical data to reliably diagnose STEMI, producing high inter-rater agreement among our expert reviewers as well as allowing reasonably precise estimates of the accuracy of administrative ICD-10 codes. However the clinical detail held in the eMR was less sufficient to diagnose NSTEMI, indicated by a lower inter-rater agreement. Efforts should be directed towards operationalising the clinical definition of NSTEMI and improving clinical record keeping to enable an accurate description of the clinical phenotype in the eMR, and thus improve reliability of the diagnosis of NSTEMI using these data sources.Article SummaryStrengths and limitations of this studyExpert chart review provided a robust evaluation of the reliability and sufficiency of data directly extracted from the EMR for the diagnosis of AMIComputational interrogation and extraction of the eMR (via SPEED-EXTRACT) allowed us to use a wide selection for inclusion in the sample on the basis of clinical data independent of ICD-10 code, enabling the capture of missed cases (i.e., uncoded AMI) and so determine estimates for the false negative rate and sensitivityResults were necessarily based on the subset of patients with sufficient clinical data in the eMR. Inferences from this subset to the wider patient pool will be biased when the availability of records varies with diagnosisAt least two sources of uncertainty in the gold reference standard we used are indistinguishable: uncertainty due to poor clinical detail in the eMR, and uncertainty due to a weak operational definition of the diagnosis (e.g., NSTEMI).

Introduction

Medical error is one of the leading causes of death, and most of these errors appear to occur in the ways that practitioners’ thoughts and feelings impact their decision making. Major gains have been made in the cognitive sciences in the past few decades that have provided a model for understanding how decisions are made—dual process theory. It is an excellent platform on which to examine the different ways decisions are made. Importantly, it allows for the examination of the pervasive influence of cognitive and affective biases on clinical decision making. Current medical training appears to fall short of what is needed to produce rational decision makers, due to what has been referred to as a mindware gap. Practitioners need to move from routine expertise to a higher level of expertise that will close this gap. A clear difficulty lies in finding ways of understanding and teaching the clinical decision-making process that do not violate the ecological characteristics of real-time clinical practice. By preserving as much as possible the rich clinical detail that makes up clinical medicine, this book attempts to offer important insights into the process.

Decreased fracture rate by mandating bone-protecting agents in the EORTC 1333/PEACE III trial comparing enzalutamide and Ra223 versus enzalutamide alone: An interim safety analysis.

5007 Background: Skeletal fractures, pathological or not, are a frequent and underestimated side-effect of systemic treatment of metastatic castration resistant prostate cancer (mCRPC). The ERA223 trial (NCT02043678) was recently unblinded following the report of a significant increase in the fracture rates when abiraterone is combined with Ra223. Hence, FDA and EMA advised against this combination. The question whether mandated use of bone protecting agents (BPA), zoledronic acid or denosumab, would have mitigated the fracture risk and whether this risk also exists in the enzalutamide/Ra223 combination is presently unknown. Methods: The phase III EORTC-1333-GUCG/PEACEIII (NCT02194842) trial compares enzalutamide vs. a combination of Ra223 and enzalutamide in asymptomatic or mildly symptomatic mCRPC patients (https://www.eortc.org/research_field/clinical-detail/1333/). After the unblinding of ERA223, the trial was amended (v4.0, April 19, 2018) to mandate that all patients must start a BPA. We report the fracture rate in the safety population of 146 treated patients as of 28/01/2019. Results: Overall, 54.2% of the patients in the enza/Ra223 arm and 51.4% of the enza arm did not receive BPA; 18.0% in the enza/Ra223 arm and 27.0% in the enza arm did not use BPA at randomization, but started during protocol treatment according to the v4.0 amendment. 27.8% and 21.6% respectively, received BPA as of randomization. In total, 45.8% of enza/Ra223 patients and 48.6% of enza only patients receive bone protection on treatment. The fracture rate is reported in the table. Conclusions: There is a 13% risk of fracture with enzalutamide in asymptomatic mCRPC, in line with previous reports. This risk is significantly increased to 33% when Ra223 is added to enzalutamide. Strikingly, the risk is almost abolished by mandatory continuous administration of BPA starting at least 6 weeks before the first injection of Ra223, thus emphasizing the importance of treating mCRPC patients with BPA. Clinical trial information: NCT02194842. [Table: see text]

Butterfly Vertebrae: A Systematic Review of the Literature and Analysis

Study Design: Systematic review (Level 4) Objective: To summarize the demographics, clinical presentations, and conditions associated with butterfly vertebrae. Methods: A systematic search was performed of multiple databases. A total of 279 articles were identified for screening. Case series or case reports of butterfly vertebrae with adequate clinical detail were complied. Results: Eighty-two total articles (109 patients) were selected for final inclusion. Sixty-one percent of patients presented with a single butterfly vertebra, while 39% were multiple. The most common location for butterfly vertebrae was T1. Fifty-six percent of cases were associated with a syndrome, the most common being spondylocostal dysostosis. The presence of multiple butterfly vertebra was strongly associated with a syndrome or additional anomalies ( P < .001). Overall, the most common presenting complaint was low back pain. Seventy percent of patients had associated spinal disease. Other organ systems affected included musculoskeletal (43%), craniofacial (30%), neurologic (27%), cardiovascular (24%), genitourinary (23%), gastrointestinal (22%), laboratory abnormality (16%), and endocrine (9%). Conclusions: This study is the largest collection of butterfly vertebrae cases to date. Butterfly vertebrae are associated with spinal deformity and multiple butterfly vertebrae may indicate a syndromic illness. Low back pain or disc herniation may occur with lumbar butterfly vertebrae however the etiology of this phenomena has not been rigorously explained. Many diseases and syndromes are associated with butterfly vertebrae.

Telemedicine in long-term care of glaucoma patients

Introduction This manuscript describes data from an original study, simulating a tele-glaucoma programme in an established clinic practice with an interdisciplinary team. This is a ‘real life’ trial of a telemedicine approach to see a follow-up patient. The goal is to evaluate the accuracy of such a programme to detect worsening and/or unstable disease. Such a programme is attractive since in-clinic time could be reduced for both the patient and provider. This study evaluates agreement between in-person and remote assessment of glaucoma progression. Methods A total of 200 adult glaucoma patients were enrolled at a single institution. The in-person assessment by an optometrist or glaucoma specialist at time of enrolment was used as the gold standard for defining progression. Collated clinical data were then reviewed by four masked providers who classified glaucoma as progression or non-progression in each eye by comparing data from enrolment visit to data from the visit immediately prior to enrolment. Agreement of glaucoma progression between the masked observer and the in-person assessment was determined using Kappa statistics. Intra-observer agreement was calculated using Kappa to compare in-person to remote assessment when both assessments were performed by the same provider ( n = 279 eyes). Results A total of 399 eyes in 200 subjects were analysed. Agreement between in-person versus remote assessment for the determination of glaucoma progression was 63%, 62%, 69% and 68% for each reader 1–4 (kappa values = 0.19, 0.20, 0.35 and 0.33, respectively). For intra-observer agreement, reader 1 agreed with their own in-person assessment for 65% of visits (kappa = 0.18). Discussion Intra-observer agreement was similar to the agreement for each provider who did not see the patient in person. This similarity suggests that telemedicine may be equally effective at identifying glaucomatous disease progression, regardless of whether the same provider performed both in-clinic and remote assessments. However, fair agreement levels highlight a limitation of using only telemedicine data to determine progression compared with clinical detail available during in-patient assessment.

Linking primary care EMR data and administrative data in Alberta, Canada: experiences, challenges, and potential solutions

IntroductionAdministrative data are commonly used for a variety of secondary purposes. Although they lack clinical detail and risk factor information, linkage to primary care electronic medical records (EMR) could fill this gap. Primary care EMRs are a relatively new data source available in Alberta and thus, EMR-administrative linkages are novel. Objectives and ApproachTo describe the process undertaken for linking de-identified primary care EMR data from two regional Alberta networks of the Canadian Primary Care Sentinel Surveillance Network (CPCSSN) with administrative data (hospital admissions, emergency department visits, pharmacy information) from Alberta Health Services Analytics, specifically as it relates to a study on patients with complex, chronic diseases. As this linkage process is new in Alberta, we will describe the challenges encountered and possible solutions to inform future data linkage for research studies. ResultsLinkage steps: 1) approval from research ethics board and individual CPCSSN providers as data custodians; 2) notify Privacy Commissioner on behalf of custodian; 3) send linking key (CPCSSN patient ID, EMR ID) from regional database to Analytics; 4) send linking files (patient personal health number [PHN], EMR ID) from custodian’s EMR system to Analytics; 5) match unique EMR ID from linking key and clinic linking files; 6) PHN from clinic linking file mapped to administrative data; 7) data de-identified before transferring to secure repository; administrative data matched to EMR data using CPCSSN ID. Challenges: obtaining individual provider consent for each study; sampling bias; delays/issues generating clinic linkage file; mismatch between patients in clinic \& regional linking files. Current and potential solutions will be discussed during the presentation. Conclusion/ImplicationsAs primary care EMR and administrative data become more routinely linked and accepted, the process will become more efficient and streamlined. These data will contribute to a better understanding of patients and their care in Alberta.

Enhancing description of hospital-conditions with ICD-11 cluster coding: Better codes for monitoring and prevention

IntroductionExposure to health care events sometimes has unintended and undesired consequences. Health care and complications arising in the course of care are diverse and complex. Representing them comprehensively in information systems is challenging, and presently beyond the bounds of practicality for routine administrative information systems that include ICD coded data. Objectives and ApproachThe ICD-11 conceptual model for hospital-acquired conditions has 3 components: 1) harm to patient 2) cause or source of harm and 3) mode or mechanism. A key feature of the Quality and Safety (Q\&S) code-set in ICD-11 is that a cluster of codes is required to represent an event or injury. Use of the term ‘cluster’ is novel in ICD-11 and so is the extent and the requirement for post-coordination. The cluster required to code a Q\&S case has three codes, one for each of the three components of the model given above. ResultsThe first component, ‘harm’, is represented by an ICD–11 diagnosis code, from any chapter of the classification. Q\&S causes or sources of harm fall into 4 types that capture events caused by substances (drugs and medicaments, etc.), procedures, devices, and a mix of other types of causes (e.g. problems associated with transfusions, incorrect diagnosis, etc.). Q\&S ‘mode or mechanism’ refers to the main way in which the ‘cause’ leads to the ‘harm’ and are specific to the type of ‘cause’ (Table 1). Table 1 - Examples of corresponding Q\&S Mode or Mechanism Cause or Source of Harm Mode or Mechanism Substance Overdose, under-dose, wrong substance. Procedure Accidental perforation of an organ during a procedure. Device Dislodgement. Malfunction. Other cause Mismatched blood. Patient dropped during transfer from OR table. Conclusion/ImplicationsThis new conceptual model for coding healthcare-related harm, dependent on the clustering of codes, has great potential to improve the clinical detail of adverse event descriptions, and the overall quality of coded health data, for better monitoring and strategies for prevention.

Abstract 82: Patient and Hospital Characteristics of Pregnancy-Related Stroke from the Get With The Guidelines Stroke Registry

Background: Stroke accounts for 14% of maternal deaths. Our knowledge of the risk factors and etiologies of pregnancy-related stroke (PRS) is limited, as most data are derived from small, single center series or large, administrative datasets lacking clinical detail. We sought to describe the patient and hospital characteristics of PRS by analyzing the Get with the Guidelines (GWTG) Stroke Registry. Methods: All female patients aged 18-44 entered into GWTG from 2008-2013 with PRS were ascertained by medical history of pregnancy (i.e. pregnant or <6 weeks postpartum) plus a principal diagnosis ICD-9 code (430, 431) (58%), PRS ICD-9 code (671.5x, 673.04, 674.0x) as the principal diagnosis alone (18%), or with a medical history of pregnancy (24%). Proportions for categorical and medians for continuous variables are reported. Results: We identified 46043 patients with stroke from 1554 sites, of whom 668 (1.5%) had PRS. Ischemic stroke (IS) occurred in 338 (51%), intracerebral hemorrhage (ICH) in 178 (27%) and subarachnoid hemorrhage (SAH) in 152 (23%). Many patient and hospital characteristics differed significantly by stroke subtype (Table). Hypertension, smoking and pre-stroke therapy with antithrombotics or antihypertensives were common; 7.4% of IS were recurrent. About 86% of all strokes did not occur in a healthcare setting and only 27% of patients arrived by EMS. Median initial blood pressure (BP) was higher in HS (ICH and SAH) than in IS, and half of all patients had initial BP below the threshold for pre-eclampsia (140/90 mmHg). HS patients were more often treated at larger, academic hospitals. Conclusions: PRS constituted 1.5% of all strokes aged 18-44 in a large contemporary stroke registry and 50% were HS. Most PRS occurred out of hospital, and half of all cases presented with normal BP levels. Further research is needed to better define PRS etiology.