Automatic ICD-10 coding algorithm using an improved longest common subsequence based on semantic similarity.
Ontology highlight
ABSTRACT:
SUBMITTER: Chen Y
PROVIDER: S-EPMC5356997 | biostudies-literature |
REPOSITORIES: biostudies-literature
Similar Datasets
Project description:BackgroundThe International Classification of Diseases, 10th Revision (ICD-10) has been widely used to describe the diagnosis information of patients. Automatic ICD-10 coding is important because manually assigning codes is expensive, time consuming and error prone. Although numerous approaches have been developed to explore automatic coding, few of them have been applied in practice. Our aim is to construct a practical, automatic ICD-10 coding machine to improve coding efficiency and quality in daily work.MethodsIn this study, we propose the use of regular expressions (regexps) to establish a correspondence between diagnosis codes and diagnosis descriptions in outpatient settings and at admission and discharge. The description models of the regexps were embedded in our upgraded coding system, which queries a diagnosis description and assigns a unique diagnosis code. Like most studies, the precision (P), recall (R), F-measure (F) and overall accuracy (A) were used to evaluate the system performance. Our study had two stages. The datasets were obtained from the diagnosis information on the homepage of the discharge medical record. The testing sets were from October 1, 2017 to April 30, 2018 and from July 1, 2018 to January 31, 2019.ResultsThe values of P were 89.27 and 88.38% in the first testing phase and the second testing phase, respectively, which demonstrate high precision. The automatic ICD-10 coding system completed more than 160,000 codes in 16 months, which reduced the workload of the coders. In addition, a comparison between the amount of time needed for manual coding and automatic coding indicated the effectiveness of the system-the time needed for automatic coding takes nearly 100 times less than manual coding.ConclusionsOur automatic coding system is well suited for the coding task. Further studies are warranted to perfect the description models of the regexps and to develop synthetic approaches to improve system performance.
| S-EPMC7157985 | biostudies-literature
Project description:Measuring the semantic similarity between Gene Ontology (GO) terms is an essential step in functional bioinformatics research. We implemented a software named GOGO for calculating the semantic similarity between GO terms. GOGO has the advantages of both information-content-based and hybrid methods, such as Resnik's and Wang's methods. Moreover, GOGO is relatively fast and does not need to calculate information content (IC) from a large gene annotation corpus but still has the advantage of using IC. This is achieved by considering the number of children nodes in the GO directed acyclic graphs when calculating the semantic contribution of an ancestor node giving to its descendent nodes. GOGO can calculate functional similarities between genes and then cluster genes based on their functional similarities. Evaluations performed on multiple pathways retrieved from the saccharomyces genome database (SGD) show that GOGO can accurately and robustly cluster genes based on functional similarities. We release GOGO as a web server and also as a stand-alone tool, which allows convenient execution of the tool for a small number of GO terms or integration of the tool into bioinformatics pipelines for large-scale calculations. GOGO can be freely accessed or downloaded from http://dna.cs.miami.edu/GOGO/ .
| S-EPMC6180005 | biostudies-literature
Project description:With the increasing amount of data made available in the chemical field, there is a strong need for systems capable of comparing and classifying chemical compounds in an efficient and effective way. The best approaches existing today are based on the structure-activity relationship premise, which states that biological activity of a molecule is strongly related to its structural or physicochemical properties. This work presents a novel approach to the automatic classification of chemical compounds by integrating semantic similarity with existing structural comparison methods. Our approach was assessed based on the Matthews Correlation Coefficient for the prediction, and achieved values of 0.810 when used as a prediction of blood-brain barrier permeability, 0.694 for P-glycoprotein substrate, and 0.673 for estrogen receptor binding activity. These results expose a significant improvement over the currently existing methods, whose best performances were 0.628, 0.591, and 0.647 respectively. It was demonstrated that the integration of semantic similarity is a feasible and effective way to improve existing chemical compound classification systems. Among other possible uses, this tool helps the study of the evolution of metabolic pathways, the study of the correlation of metabolic networks with properties of those networks, or the improvement of ontologies that represent chemical information.
| S-EPMC2944781 | biostudies-literature
Project description:BackgroundWhile potentially timesaving, there is no program to automatically transform diagnosis codes of the ICD-10 German modification (ICD-10-GM) into the injury severity score (ISS).ObjectiveTo develop a mapping method from ICD-10-GM into ICD-10 clinical modification (ICD-10-CM) to calculate the abbreviated injury scale (AIS) and ISS of each patient using the ICDPIC-R and to compare the manually and automatically calculated scores.MethodsBetween January 2019 and June 2021, the most severe AIS of each body region and the ISS were manually calculated using medical documentation and radiology reports of all major trauma patients of a German level I trauma centre. The ICD-10-GM codes of these patients were exported from the electronic medical data system SAP, and a Java program was written to transform these into ICD-10-CM codes. Afterwards, the ICDPIC-R was used to automatically generate the most severe AIS of each body region and the ISS. The automatically and manually determined ISS and AIS scores were then tested for equivalence.ResultsStatistical analysis revealed that the manually and automatically calculated ISS were significantly equivalent over the entire patient cohort. Further sub-group analysis, however, showed that equivalence could only be demonstrated for patients with an ISS between 16 and 24. Likewise, the highest AIS scores of each body region were not equal in the manually and automatically calculated group.ConclusionThough achieving mapping results highly comparable to previous mapping methods of ICD-10-CM diagnosis codes, it is not unrestrictedly possible to automatically calculate the AIS and ISS using ICD-10-GM codes.
| S-EPMC8432850 | biostudies-literature
Project description:The revision of the International Classification of Diseases (ICD) could change morbidity and mortality statistics significantly, which also affects the area of infectious diseases. Infectious diseases are classified according to their etiology, affected body system or the life period during which the episode occurs. Specific challenges arise from emerging pathogens and the respective necessary adaptation. For epidemiologic analysis ICD-10 does not always offer enough additional information.ICD provides the basis for international comparison of infectious disease morbidity and mortality statistics, but it is also used to collect data for surveillance and research purposes, e.?g. the notification system for infectious diseases, syndromic surveillance systems and the evaluation of data quality by using secondary data sources.ICD-11 offers the chance to better represent epidemiological concepts of infectious diseases by adding more relevant information as affected body system or manifestation. Due to the complexity of coding, ensuring continuity of morbidity and mortality statistics could be challenging.
| S-EPMC7079900 | biostudies-literature
Project description:ObjectiveWe describe the most frequently used musculoskeletal diagnoses in Veterans Health Administration care. We report the number of visits and patients associated with common musculoskeletal International Classification of Diseases (ICD)-10 codes and compare trends across primary and specialty care settings.DesignSecondary analysis of a longitudinal cohort study.SubjectsVeterans included in the Musculoskeletal Diagnosis Cohort with a musculoskeletal diagnosis from October 1, 2015, through September 30, 2017.MethodsWe obtained counts and proportions of all musculoskeletal diagnosis codes used and the number of unique patients with each musculoskeletal diagnosis. Diagnosis use was compared between primary and specialty care settings.ResultsOf more than 6,400 possible ICD-10 M-codes describing "Diseases of the Musculoskeletal System and Connective Tissue," 5,723 codes were used at least once. The most frequently used ICD-10 M-code was "Low Back Pain" (18.3%), followed by "Cervicalgia" (3.6%). Collectively, the 100 most frequently used codes accounted for 80% of M-coded visit diagnoses, and 95% of patients had at least one of these diagnoses. The most common diagnoses (spinal pain, joint pain, osteoarthritis) were used similarly in primary and specialty care settings.ConclusionA diverse sample of all available musculoskeletal diagnosis codes were used; however, less than 2% of all possible codes accounted for 80% of the diagnoses used. This trend was consistent across primary and specialty care settings. The most frequently used diagnosis codes describe the types of musculoskeletal conditions, among a large pool of potential diagnoses, that prompt veterans to present to the Veterans Health Administration for musculoskeletal care.
| S-EPMC8783617 | biostudies-literature
Project description:Genetic maps order genetic markers along chromosomes. They are, for instance, extensively used in marker-assisted selection to accelerate breeding programs. Even for the same species, people often have to deal with several alternative maps obtained using different ordering methods or different datasets, e.g. resulting from different segregating populations. Having efficient tools to identify the consistency and discrepancy of alternative maps is thus essential to facilitate genetic map comparisons. We propose to encode genetic maps by bucket order, a kind of order, which takes into account the blurred parts of the marker order while being an efficient data structure to achieve low complexity algorithms. The main result of this paper is an O(n log(n)) procedure to identify the largest agreements between two bucket orders of n elements, their Longest Common Subsequence (LCS), providing an efficient solution to highlight discrepancies between two genetic maps. The LCS of two maps, being the largest set of their collinear markers, is used as a building block to compute pairwise map congruence, to visually emphasize maker collinearity and in some scaffolding methods relying on genetic maps to improve genome assembly. As the LCS computation is a key subroutine of all these genetic map related tools, replacing the current LCS subroutine of those methods by ours -to do the exact same work but faster- could significantly speed up those methods without changing their accuracy. To ease such transition we provide all required algorithmic details in this self contained paper as well as an R package implementing them, named LCSLCIS, which is freely available at: https://github.com/holtzy/LCSLCIS.
| S-EPMC6320017 | biostudies-literature
Project description:PurposeLymphoma is a health outcome of interest for drug safety studies. Studies using administrative claims data require the accurate identification of lymphoma cases. We developed and validated an International Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM)-based algorithm to identify lymphoma in healthcare claims data.MethodsWe developed a three-component algorithm to identify patients aged ≥15 years who were newly diagnosed with Hodgkin (HL) or non-Hodgkin (NHL) lymphoma from January 2016 through July 2018 among members of four Data Partners within the FDA's Sentinel System. The algorithm identified potential cases as patients with ≥2 ICD-10-CM lymphoma diagnosis codes on different dates within 183 days; ≥1 procedure code for a diagnostic procedure (e.g., biopsy, flow cytometry) and ≥1 procedure code for a relevant imaging study within 90 days of the first lymphoma diagnosis code. Cases identified by the algorithm were adjudicated via chart review and a positive predictive value (PPV) was calculated.ResultsWe identified 8723 potential lymphoma cases via the algorithm and randomly sampled 213 for validation. We retrieved 138 charts (65%) and adjudicated 134 (63%). The overall PPV was 77% (95% confidence interval: 69%-84%). Most cases also had subtype information available, with 88% of cases identified as NHL and 11% as HL.ConclusionsSeventy-seven percent of lymphoma cases identified by an algorithm based on ICD-10-CM diagnosis and procedure codes and applied to claims data were true cases. This novel algorithm represents an efficient, cost-effective way to target an important health outcome of interest for large-scale drug safety and public health surveillance studies.
| S-EPMC8205565 | biostudies-literature
Project description:BackgroundWe aimed to estimate the impact of the International Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM) coding transition on traumatic injury-related hospitalization trends among young adults across a geographically and demographically diverse group of U.S. states.MethodsInterrupted time series analyses were conducted using statewide inpatient databases from 12 states and including traumatic injury-related hospitalizations in adults aged 19-44?years in 2011-2017. Segmented regression models were used to estimate the impact of the October 2015 coding transition on external cause of injury (ECOI) completeness (percentage of hospitalizations with a documented ECOI code) and on population-level rates of injury-related hospitalizations by nature, intent, mechanism, and severity of injury.ResultsThe transition to ICD-10-CM was associated with a drop in ECOI completion in the transition month (-?3.7%; P?<?.0001), but there was no significant change in the positive trend in ECOI completion from the pre- to post-transition periods. There were significant increases post-transition in the measured rates of hospitalization for traumatic brain injury (TBI), unintentional injury, mild injury (injury severity score (ISS) <?9), and injuries caused by drowning, firearms, machinery, other pedestrian, suffocation, and unspecified mechanism. Conversely, there were significant decreases in October 2015 in the rates of hospitalization for assault, injuries of undetermined intent, injuries of moderate severity (ISS 9-15), and injuries caused by fire/burn, other pedal cyclist, other transportation, natural/environmental, and other specified mechanism. A significant increase in the percentage of hospitalizations classified as resulting from severe injury (ISS?>?15) was observed when the general equivalence mapping maximum severity method for converting ICD-10-CM codes to ICD-9-CM codes was used. State-specific results for the outcomes of ECOI completion and TBI-related hospitalization rates are provided in an online supplement.ConclusionsThe U.S. transition from ICD-9-CM to ICD-10-CM coding led to a significant decrease in ECOI completion and several significant changes in measured rates of injury-related hospitalizations by injury intent, mechanism, nature, and severity. The results of this study can inform the design and analysis of future traumatic injury-related health services research studies that use both ICD-9-CM and ICD-10-CM coded data.Level of evidenceII (Interrupted Time Series).
| S-EPMC7830822 | biostudies-literature
Project description:Semantic similarity measures are useful to assess the physiological relevance of protein-protein interactions (PPIs). They quantify similarity between proteins based on their function using annotation systems like the Gene Ontology (GO). Proteins that interact in the cell are likely to be in similar locations or involved in similar biological processes compared to proteins that do not interact. Thus the more semantically similar the gene function annotations are among the interacting proteins, more likely the interaction is physiologically relevant. However, most semantic similarity measures used for PPI confidence assessment do not consider the unequal depth of term hierarchies in different classes of cellular location, molecular function, and biological process ontologies of GO and thus may over-or under-estimate similarity.We describe an improved algorithm, Topological Clustering Semantic Similarity (TCSS), to compute semantic similarity between GO terms annotated to proteins in interaction datasets. Our algorithm, considers unequal depth of biological knowledge representation in different branches of the GO graph. The central idea is to divide the GO graph into sub-graphs and score PPIs higher if participating proteins belong to the same sub-graph as compared to if they belong to different sub-graphs.The TCSS algorithm performs better than other semantic similarity measurement techniques that we evaluated in terms of their performance on distinguishing true from false protein interactions, and correlation with gene expression and protein families. We show an average improvement of 4.6 times the F1 score over Resnik, the next best method, on our Saccharomyces cerevisiae PPI dataset and 2 times on our Homo sapiens PPI dataset using cellular component, biological process and molecular function GO annotations.
| S-EPMC2998529 | biostudies-literature