Phenotype execution and modeling architecture to support disease surveillance and real-world evidence studies: English sentinel network evaluation.

Objective: To evaluate Phenotype Execution and Modelling Architecture (PhEMA), to express sharable phenotypes using Clinical Quality Language (CQL) and intensional Systematised Nomenclature of Medicine (SNOMED) Clinical Terms (CT) Fast Healthcare Interoperability Resources (FHIR) valuesets, for exemplar chronic disease, sociodemographic risk factor, and surveillance phenotypes. Method: We curated 3 phenotypes: Type 2 diabetes mellitus (T2DM), excessive alcohol use, and incident influenza-like illness (ILI) using CQL to define clinical and administrative logic. We defined our phenotypes with valuesets, using SNOMED's hierarchy and expression constraint language, and CQL, combining valuesets and adding temporal elements where needed. We compared the count of cases found using PhEMA with our existing approach using convenience datasets. We assessed our new approach against published desiderata for phenotypes. Results: The T2DM phenotype could be defined as 2 intensionally defined SNOMED valuesets and a CQL script. It increased the prevalence from 7.2% to 7.3%. Excess alcohol phenotype was defined by valuesets that added qualitative clinical terms to the quantitative conceptual definitions we currently use; this change increased prevalence by 58%, from 1.2% to 1.9%. We created an ILI valueset with SNOMED concepts, adding a temporal element using CQL to differentiate new episodes. This increased the weekly incidence in our convenience sample (weeks 26-38) from 0.95 cases to 1.11 cases per 100 000 people. Conclusions: Phenotypes for surveillance and research can be described fully and comprehensibly using CQL and intensional FHIR valuesets. Our use case phenotypes identified a greater number of cases, whilst anticipated from excessive alcohol this was not for our other variable. This may have been due to our use of SNOMED CT hierarchy. Our new process fulfilled a greater number of phenotype desiderata than the one that we had used previously, mostly in the modeling domain. More work is needed to implement that sharing and warehousing domains.

Targeted education for clinicians and clinical coding staff improves the accuracy of clinical coding: A quality improvement project.

The allocation of healthcare resources is reliant upon accurate information generated through clinical coding. Several factors contribute to coding inaccuracies, one of which is interpreting medical documentation. A lack of awareness among medical staff of the clinical coding process and the importance of detailed documentation exacerbates this problem. To investigate this further, 1 month of inpatient clinical coding data from a single hospital ward was reviewed by clinicians experienced in the coding and auditing process. If the reviewing clinician identified inaccuracies in the initial clinical coding, Healthcare Resource Group (HRG) codes were changed. Education sessions were then provided both to junior clinicians working on the hospital ward and to clinical coding staff and a further month of clinical coding data was again reviewed to assess for any difference after the sessions. HRG changes were made in 58.5% of 94 cases initially. Following the educational sessions, 20.5% of HRGs changed in 73 cases (p<0.0001), indicating more accurate initial clinical coding. There were also statistically significant reductions in the extent to which the primary and secondary diagnoses were changed. This study demonstrates that targeted education sessions for both junior clinicians and clinical coding staff can improve the accuracy of inpatient clinical coding.

Prevalence and risk factors for long COVID among adults in Scotland using electronic health records: a national, retrospective, observational cohort study.

Background: Long COVID is a debilitating multisystem condition. The objective of this study was to estimate the prevalence of long COVID in the adult population of Scotland, and to identify risk factors associated with its development. Methods: In this national, retrospective, observational cohort study, we analysed electronic health records (EHRs) for all adults (≥18 years) registered with a general medical practice and resident in Scotland between March 1, 2020, and October 26, 2022 (98-99% of the population). We linked data from primary care, secondary care, laboratory testing and prescribing. Four outcome measures were used to identify long COVID: clinical codes, free text in primary care records, free text on sick notes, and a novel operational definition. The operational definition was developed using Poisson regression to identify clinical encounters indicative of long COVID from a sample of negative and positive COVID-19 cases matched on time-varying propensity to test positive for SARS-CoV-2. Possible risk factors for long COVID were identified by stratifying descriptive statistics by long COVID status. Findings: Of 4,676,390 participants, 81,219 (1.7%) were identified as having long COVID. Clinical codes identified the fewest cases (n = 1,092, 0.02%), followed by free text (n = 8,368, 0.2%), sick notes (n = 14,469, 0.3%), and the operational definition (n = 64,193, 1.4%). There was limited overlap in cases identified by the measures; however, temporal trends and patient characteristics were consistent across measures. Compared with the general population, a higher proportion of people with long COVID were female (65.1% versus 50.4%), aged 38-67 (63.7% versus 48.9%), overweight or obese (45.7% versus 29.4%), had one or more comorbidities (52.7% versus 36.0%), were immunosuppressed (6.9% versus 3.2%), shielding (7.9% versus 3.4%), or hospitalised within 28 days of testing positive (8.8% versus 3.3%%), and had tested positive before Omicron became the dominant variant (44.9% versus 35.9%). The operational definition identified long COVID cases with combinations of clinical encounters (from four symptoms, six investigation types, and seven management strategies) recorded in EHRs within 4-26 weeks of a positive SARS-CoV-2 test. These combinations were significantly (p < 0.0001) more prevalent in positive COVID-19 patients than in matched negative controls. In a case-crossover analysis, 16.4% of those identified by the operational definition had similar healthcare patterns recorded before testing positive. Interpretation: The prevalence of long COVID presenting in general practice was estimated to be 0.02-1.7%, depending on the measure used. Due to challenges in diagnosing long COVID and inconsistent recording of information in EHRs, the true prevalence of long COVID is likely to be higher. The operational definition provided a novel approach but relied on a restricted set of symptoms and may misclassify individuals with pre-existing health conditions. Further research is needed to refine and validate this approach. Funding: Chief Scientist Office (Scotland), Medical Research Council, and BREATHE.

Validity of ICD-10 diagnosis codes for placenta accreta spectrum disorders.

BACKGROUND: The 10th revision of the International Classification of Diseases, Clinical Modification (ICD-10) includes diagnosis codes for placenta accreta spectrum for the first time. These codes could enable valuable research and surveillance of placenta accreta spectrum, a life-threatening pregnancy complication that is increasing in incidence. OBJECTIVE: We sought to evaluate the validity of placenta accreta spectrum diagnosis codes that were introduced in ICD-10 and assess contributing factors to incorrect code assignments. METHODS: We calculated sensitivity, specificity, positive predictive value and negative predictive value of the ICD-10 placenta accreta spectrum code assignments after reviewing medical records from October 2015 to March 2020 at a quaternary obstetric centre. Histopathologic diagnosis was considered the gold standard. RESULTS: Among 22,345 patients, 104 (0.46%) had an ICD-10 code for placenta accreta spectrum and 51 (0.23%) had a histopathologic diagnosis. ICD-10 codes had a sensitivity of 0.71 (95% CI 0.56, 0.83), specificity of 0.98 (95% CI 0.93, 1.00), positive predictive value of 0.61 (95% CI 0.48, 0.72) and negative predictive value of 1.00 (95% CI 0.96, 1.00). The sensitivities of the ICD-10 codes for placenta accreta spectrum subtypes- accreta, increta and percreta-were 0.55 (95% CI 0.31, 0.78), 0.33 (95% CI 0.12, 0.62) and 0.56 (95% CI 0.31, 0.78), respectively. Cases with incorrect code assignment were less morbid than cases with correct code assignment, with a lower incidence of hysterectomy at delivery (17% vs 100%), blood transfusion (26% vs 75%) and admission to the intensive care unit (0% vs 53%). Primary reasons for code misassignment included code assigned to cases of occult placenta accreta (35%) or to cases with clinical evidence of placental adherence without histopatholic diagnostic (35%) features. CONCLUSION: These findings from a quaternary obstetric centre suggest that ICD-10 codes may be useful for research and surveillance of placenta accreta spectrum, but researchers should be aware of likely substantial false positive cases.

Improving Quality of ICD-10 (International Statistical Classification of Diseases, Tenth Revision) Coding Using AI: Protocol for a Crossover Randomized Controlled Trial.

BACKGROUND: Computer-assisted clinical coding (CAC) tools are designed to help clinical coders assign standardized codes, such as the ICD-10 (International Statistical Classification of Diseases, Tenth Revision), to clinical texts, such as discharge summaries. Maintaining the integrity of these standardized codes is important both for the functioning of health systems and for ensuring data used for secondary purposes are of high quality. Clinical coding is an error-prone cumbersome task, and the complexity of modern classification systems such as the ICD-11 (International Classification of Diseases, Eleventh Revision) presents significant barriers to implementation. To date, there have only been a few user studies; therefore, our understanding is still limited regarding the role CAC systems can play in reducing the burden of coding and improving the overall quality of coding. OBJECTIVE: The objective of the user study is to generate both qualitative and quantitative data for measuring the usefulness of a CAC system, Easy-ICD, that was developed for recommending ICD-10 codes. Specifically, our goal is to assess whether our tool can reduce the burden on clinical coders and also improve coding quality. METHODS: The user study is based on a crossover randomized controlled trial study design, where we measure the performance of clinical coders when they use our CAC tool versus when they do not. Performance is measured by the time it takes them to assign codes to both simple and complex clinical texts as well as the coding quality, that is, the accuracy of code assignment. RESULTS: We expect the study to provide us with a measurement of the effectiveness of the CAC system compared to manual coding processes, both in terms of time use and coding quality. Positive outcomes from this study will imply that CAC tools hold the potential to reduce the burden on health care staff and will have major implications for the adoption of artificial intelligence-based CAC innovations to improve coding practice. Expected results to be published summer 2024. CONCLUSIONS: The planned user study promises a greater understanding of the impact CAC systems might have on clinical coding in real-life settings, especially with regard to coding time and quality. Further, the study may add new insights on how to meaningfully exploit current clinical text mining capabilities, with a view to reducing the burden on clinical coders, thus lowering the barriers and paving a more sustainable path to the adoption of modern coding systems, such as the new ICD-11. TRIAL REGISTRATION: clinicaltrials.gov NCT06286865; https://clinicaltrials.gov/study/NCT06286865. INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID): DERR1-10.2196/54593.

Use of Healthcare Claims Data to Generate Real-World Evidence on Patients With Drug-Resistant Epilepsy: Practical Considerations for Research.

Objectives: Regulatory bodies, health technology assessment agencies, payers, physicians, and other decision-makers increasingly recognize the importance of real-world evidence (RWE) to provide important and relevant insights on treatment patterns, burden/cost of illness, product safety, and long-term and comparative effectiveness. However, RWE generation requires a careful approach to ensure rigorous analysis and interpretation. There are limited examples of comprehensive methodology for the generation of RWE on patients who have undergone neuromodulation for drug-resistant epilepsy (DRE). This is likely due, at least in part, to the many challenges inherent in using real-world data to define DRE, neuromodulation (including type implanted), and related outcomes of interest. We sought to provide recommendations to enable generation of robust RWE that can increase knowledge of "real-world" patients with DRE and help inform the difficult decisions regarding treatment choices and reimbursement for this particularly vulnerable population. Methods: We drew upon our collective decades of experience in RWE generation and relevant disciplines (epidemiology, health economics, and biostatistics) to describe challenges inherent to this therapeutic area and to provide potential solutions thereto within healthcare claims databases. Several examples were provided from our experiences in DRE to further illustrate our recommendations for generation of robust RWE in this therapeutic area. Results: Our recommendations focus on considerations for the selection of an appropriate data source, development of a study timeline, exposure allotment (specifically, neuromodulation implantation for patients with DRE), and ascertainment of relevant outcomes. Conclusions: The need for RWE to inform healthcare decisions has never been greater and continues to grow in importance to regulators, payers, physicians, and other key stakeholders. However, as real-world data sources used to generate RWE are typically generated for reasons other than research, rigorous methodology is required to minimize bias and fully unlock their value.

Challenges and successes in implementing an integrated electronic patient record (HIVE) at the Manchester University National Health Service Foundation Trust, England: 1000+ legacy systems, 10 hospitals, one electronic patient record.

BACKGROUND: The Manchester University National Health Service (NHS) Foundation Trust (MFT) is one of the largest NHS Trusts in England. Historically, the Trust has had very mixed clinical record keeping, including over 1000 individual information systems. None of these health information technology systems had the full functionality of an integrated electronic patient record (EPR). MFT evolved to its current size and complexity with a vision to improve patient care in Greater Manchester by adopting a Trust-wide EPR. The EPR "Go Live" occurred in September 2022. AIM: To describe the process of EPR integration as it reflected and impacted upon MFT's health information management (HIM) teams. METHOD: MFT worked through a 2-year readiness program of work. This included technical readiness, software development and migration planning. Migration of data from the approximately 1000 systems was a major undertaking, during which access to the clinical history and ongoing operational reporting needed to be maintained. Pre-implementation requirements were outlined, a change management program was implemented, and the overall implementation was managed to tight timelines. DISCUSSION: "Go Live" was achieved for the EPIC EPR product (HIVE) within MFT. Legacy systems are still in the process of being decommissioned and staff are transacting within HIVE. Significant changes in processes and reporting continue to be made, despite some challenges. CONCLUSION: The Trust delivered the single largest EPIC European "Go live." Lessons learnt continue to be identified. The impact of what the EPR means for the HIM function is described.

Development of a New Knee Arthroscopy Operative Proforma Saving Thousands of British Pounds.

Coding inaccuracies in documentation of surgical procedures misrepresent the productivity of departments, with harmful fiscal consequences and detract from effective clinical governance. We aimed to assess the extent of this within our centres.  We retrospectively analysed the operative records of 34 patients from two centres over a period of a month, undergoing varying arthroscopic knee operations. We found that 50% of cases had incorrect coding for procedures performed. On review of the clinical coding, the loss of payment summed up to £29,325.  The flawed coding practices stemmed from the heterogeneity and convolution in documentation of procedures. Our intervention was the development of a multi-faceted arthroscopic operation note proforma, centred on concise documentation for appropriate codes to be gleaned. We re-audited our new proforma, retrospectively collating data on 37 patients over a period of five months undergoing arthroscopic knee procedures. We found only 5% of cases were coded incorrectly, summing to a loss in tariff payment of £2654.  In conclusion, poor quality of documentation and written communication between surgical and coding departments can have drastic ramifications for funding. An active refinement of this process can ultimately help to provide more resources for improved patient care.

Sepsis-2.5: Resolving Conflicts Between Payers and Providers.

Competing definitions of sepsis have significant clinical implications and impact both medical coding and hospital payment. Although clinicians may prefer Sepsis-2, payer use of Sepsis-3 to validate clinical diagnoses may result in denial of payment or requests to recoup previously paid funds from healthcare providers. The Sepsis-2.5 project was a cooperative effort between a hospital system and a private payer to develop a community-based, literature-supported consensus definition for sepsis characterized by the presence of clinical illness, a source of infection, and evidence of organ dysfunction. This new definition ("Sepsis-2.5") has been instrumental in resolving provider-payer conflicts in defining clinical sepsis and reimbursing care.

Population sepsis incidence, mortality, and trends in Hong Kong between 2009-2018 using clinical and administrative data.

BACKGROUND: Sepsis surveillance using electronic health record (EHR)-based data may provide more accurate epidemiologic estimates than administrative data, but experience with this approach to estimate population-level sepsis burden is lacking. METHODS: This was a retrospective cohort study including all adults admitted to publicly-funded hospitals in Hong Kong between 2009-2018. Sepsis was defined as clinical evidence of presumed infection (clinical cultures and treatment with antibiotics) and concurrent acute organ dysfunction (≥2 point increase in baseline SOFA score). Trends in incidence, mortality, and case fatality risk (CFR) were modelled by exponential regression. Performance of the EHR-based definition was compared with 4 administrative definitions using 500 medical record reviews. RESULTS: Among 13,550,168 hospital episodes during the study period, 485,057 (3.6%) had sepsis by EHR-based criteria with 21.5% CFR. In 2018, age- and sex-adjusted standardized sepsis incidence was 759 per 100,000 (relative +2.9%/year [95%CI 2.0, 3.8%] between 2009-2018) and standardized sepsis mortality was 156 per 100,000 (relative +1.9%/year [95%CI 0.9,2.9%]). Despite decreasing CFR (relative -0.5%/year [95%CI -1.0, -0.1%]), sepsis accounted for an increasing proportion of all deaths (relative +3.9%/year [95%CI 2.9, 4.9%]). Medical record reviews demonstrated that the EHR-based definition more accurately identified sepsis than administrative definitions (AUC 0.91 vs 0.52-0.55, p < 0.001). CONCLUSIONS: An objective EHR-based surveillance definition demonstrated an increase in population-level standardized sepsis incidence and mortality in Hong Kong between 2009-2018 and was much more accurate than administrative definitions. These findings demonstrate the feasibility and advantages of an EHR-based approach for widescale sepsis surveillance.

Perceptions of Portuguese medical coders on the transition to ICD-10-CM/PCS: A national survey.

BACKGROUND: In Portugal, trained physicians undertake the clinical coding process, which serves as the basis for hospital reimbursement systems. In 2017, the classification version used for coding of diagnoses and procedures for hospital morbidity changed from the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) to the International Classification of Diseases, Tenth Revision, Clinical Modification/Procedure Coding System (ICD-10-CM/PCS). OBJECTIVE: To assess the perceptions of medical coders on the transition of the clinical coding process from ICD-9-CM to ICD-10-CM/PCS in terms of its impact on data quality, as well as the major differences, advantages, and problems they faced. METHOD: We conducted an observational study using a web-based survey submitted to medical coders in Portugal. Survey questions were based on a literature review and from previous focus group studies. RESULTS: A total of 103 responses were obtained from medical coders with experience in the two versions of the classification system (i.e. ICD-9-CM and ICD-10-CM/PCS). Of these, 82 (79.6%) medical coders preferred the latest version and 76 (73.8%) considered that ICD-10-CM/PCS guaranteed higher quality of the coded data. However, more than half of the respondents (N = 61; 59.2%) believed that more time for the coding process for each episode was needed. CONCLUSION: Quality of clinical coded data is one of the major priorities that must be ensured. According to the medical coders, the use of ICD-10-CM/PCS appeared to achieve higher quality coded data, but also increased the effort. IMPLICATIONS: According to medical coders, the change off classification systems should improve the quality of coded data. Nevertheless, the extra time invested in this process might also pose a problem in the future.

Comparison of the accuracy of inpatient morbidity coding with ICD-11 and ICD-10.

BACKGROUND: One of the challenges when transitioning from International Statistical Classification of Diseases and Related Health Problems, 10th Revision (ICD-10) to International Classification of Diseases, 11th Revision (ICD-11) is to ensure clinical coding accuracy. OBJECTIVE: To determine the accuracy of clinical coding with ICD-11 in comparison with ICD-10 and identify causes of coding errors in real clinical coding environments. METHOD: The study was conducted prospectively in two general hospitals. Medical records of discharged inpatients were coded by hospital clinical coders with both ICD-11 and ICD-10 on different days. These medical records were recoded by five mentors. Codes assigned by mentors were used as the gold standard for the evaluation of accuracy. RESULTS: The accuracy of ICD-10 and ICD-11 coding for 1578 and 2168 codes was evaluated. Coding accuracy was 89.1% and 74.2% for ICD-10 and ICD-11. In ICD-11, the lowest accuracy was observed in chapters 22 (injuries), 10 (ear) and 11 (circulatory) (51.1%, 53.8% and 62.7%, respectively). In both ICD-10 and ICD-11, the most important cause of the coding errors was clinical coders' mistakes (79.5% and 81.8% for ICD-10 and ICD-11, respectively). CONCLUSION: Accuracy of clinical coding with ICD-11 was lower relative to ICD-10. Hence, it is essential to carry out initial preparations, particularly the training of clinical coders based on their needs, as well as the necessary interventions to enhance the documentation of medical records according to ICD-11 before or simultaneous with the country-wide implementation. IMPLICATIONS: Clinical coders need complete training, especially in using extension codes and post-coordination coding. Local ICD-11 guidelines based on the needs of local users and reporting policies should be developed. Furthermore, documentation guidelines based on ICD-11 requirements should be developed.

Stroke clinical coding education program in Australia and New Zealand.

BACKGROUND: Accurate coded diagnostic data are important for epidemiological research of stroke. OBJECTIVE: To develop, implement and evaluate an online education program for improving clinical coding of stroke. METHOD: The Australia and New Zealand Stroke Coding Working Group co-developed an education program comprising eight modules: rationale for coding of stroke; understanding stroke; management of stroke; national coding standards; coding trees; good clinical documentation; coding practices; and scenarios. Clinical coders and health information managers participated in the 90-minute education program. Pre- and post-education surveys were administered to assess knowledge of stroke and coding, and to obtain feedback. Descriptive analyses were used for quantitative data, inductive thematic analysis for open-text responses, with all results triangulated. RESULTS: Of 615 participants, 404 (66%) completed both pre- and post-education assessments. Respondents had improved knowledge for 9/12 questions (p < 0.05), including knowledge of applicable coding standards, coding of intracerebral haemorrhage and the actions to take when coding stroke (all p < 0.001). Majority of respondents agreed that information was pitched at an appropriate level; education materials were well organised; presenters had adequate knowledge; and that they would recommend the session to colleagues. In qualitative evaluations, the education program was beneficial for newly trained clinical coders, or as a knowledge refresher, and respondents valued clinical information from a stroke neurologist. CONCLUSION: Our education program was associated with increased knowledge for clinical coding of stroke. To continue to address the quality of coded stroke data through improved stroke documentation, the next stage will be to adapt the educational program for clinicians.

Sensitivity of ICD coding for sepsis in children-a population-based study.

Background: International Classification of Diseases 10th edition (ICD-10) is widely used to describe the burden of disease. Aim: To describe how well ICD-10 coding captures sepsis in children admitted to the hospital with blood culture-proven bacterial or fungal infection and systemic inflammatory response syndrome. Methods: Secondary analysis of a population-based, multicenter, prospective cohort study on children with blood culture-proven sepsis of nine tertiary pediatric hospitals in Switzerland. We compared the agreement of validated study data on sepsis criteria with ICD-10 coding abstraction obtained at the participating hospitals. Results: We analyzed 998 hospital admissions of children with blood culture-proven sepsis. The sensitivity of ICD-10 coding abstraction was 60% (95%-CI 57-63) for sepsis; 35% (95%-CI 31-39) for sepsis with organ dysfunction, using an explicit abstraction strategy; and 65% (95%-CI 61-69) using an implicit abstraction strategy. For septic shock, the sensitivity of ICD-10 coding abstraction was 43% (95%-CI 37-50). Agreement of ICD-10 coding abstraction with validated study data varied by the underlying infection type and disease severity (p < 0.05). The estimated national incidence of sepsis, inferred from ICD-10 coding abstraction, was 12.5 per 100,000 children (95%-CI 11.7-13.5) and 21.0 per 100,000 children (95%-CI 19.8-22.2) using validated study data. Conclusions: In this population-based study, we found a poor representation of sepsis and sepsis with organ dysfunction by ICD-10 coding abstraction in children with blood culture-proven sepsis when compared against a prospective validated research dataset. Sepsis estimates in children based on ICD-10 coding may thus severely underestimate the true prevalence of the disease. Supplementary Information: The online version contains supplementary material available at 10.1007/s44253-023-00006-1.

A Model for Clinical Coder Satisfaction in Saudi Arabia Based on a Holistic Approach: Clinical, Professional and Organizational Dimensions.

The quality of clinical coding influences not only hospital revenue but also the quality and efficiency of healthcare services. Assessing the coders' satisfaction is essential to optimizing the quality of clinical coding. This mixed-method study used a qualitative approach to propose the study model while testing the model through a quantitative approach. The relevant variables of the satisfaction model were assessed through a survey targeting clinical coders across the country on a timely basis. Fourteen experts participated in establishing the model with three dimensions: professional, organizational, and clinical. Each dimension has its relevant variables. One hundred eighty-four clinical coders participated in phase two. 34.5% were male, 61% held a diploma, 38% had a bachelor's and above, and 49.7% worked in hospitals having fully electronic health records. We found that organizational and clinical dimensions strongly correlate with coders' satisfaction. Noticeably, the most influencing variables were the availability of coding policies and the computer-assisted coding (CAC) system. The results show that the model explains the satisfaction of the clinical coders, and organizational and clinical-related variables are crucial. Although gender-based differences exist, training (regardless of the training mode), coding policies, and the CAC system substantially influence coders' satisfaction. A significant stream of the literature supports these findings. However, attempting a holistic approach to assess coders' satisfaction and affecting coding quality is the added value of this study. Optimizing clinical coding practice requires organization-wide initiatives and policies to regulate coding practices and standards to promote the quality and timeliness of clinical documentation. Training is indispensable not only for clinical coders, but physicians also need to understand the rationale and value of clinical coding. Better utilization of the outcomes of the coding process and adopting the CAC system are significant drivers to enhance coders' satisfaction.

From coding to clinical nurse specialist: how a review of coding practice enabled hysteroscopy nurse development.

Clinical coding, the method by which departments are reimbursed for providing services to patients, is widely mispractised within the NHS. Improving clinical coding accuracy therefore offers an opportunity to increase departmental income, guide efficient resource allocation and enable staff development. The authors audited the clinical coding in outpatient hysteroscopy clinics at their institution and found that coding errors were both prevalent and correctable. By implementing simple changes in coding procedure, and without any additional administrative cost, they significantly improved coding accuracy and achieved an increase in total annual tariffs. Although not applicable in a block contract, this will become highly relevant in a restoration of the Payment by Results tariff system. Nurse development is a key objective of the NHS Long Term Plan but can be hindered by staff costs, which require departmental funding. In the authors' institution, improved clinical coding accuracy directly led to a departmental restructuring, funded the development of a new hysteroscopy nurse development and improved care delivery. Coding errors are not unique to the authors' trust, yet simple amendments led to meaningful changes. Therefore, careful auditing and implemented change are needed to raise national clinical coding standards, to enable clinical restructuring, staff development, and provide more efficient, patient-centred care.

Assessing national patterns and outcomes of pituitary surgery: is hospital administrative data good enough?

PURPOSE: Patterns of surgical care, outcomes, and quality of care can be assessed using hospital administrative databases but this requires accurate and complete data. The aim of this study was to explore whether the quality of hospital administrative data was sufficient to assess pituitary surgery practice in England. METHODS: The study analysed Hospital Episode Statistics (HES) data from April 2013 to March 2018 on all adult patients undergoing pituitary surgery in England. A series of data quality indicators examined the attribution of cases to consultants, the coding of sellar and parasellar lesions, associated endocrine and visual disorders, and surgical procedures. Differences in data quality over time and between neurosurgical units were examined. RESULTS: A total of 5613 records describing pituitary procedures were identified. Overall, 97.3% had a diagnostic code for the tumour or lesion treated, with 29.7% (n = 1669) and 17.8% (n = 1000) describing endocrine and visual disorders, respectively. There was a significant reduction from the first to the fifth year in records that only contained a pituitary tumour code (63.7%-47.0%, p < .001). The use of procedure codes that attracted the highest tariff increased over time (66.4%-82.4%, p < .001). Patterns of coding varied widely between the 24 neurosurgical units. CONCLUSION: The quality of HES data on pituitary surgery has improved over time but there is wide variation in the quality of data between neurosurgical units. Research studies and quality improvement programmes using these data need to check it is of sufficient quality to not invalidate their results.

Explainable clinical coding with in-domain adapted transformers.

BACKGROUND AND OBJECTIVE: Automatic clinical coding is a crucial task in the process of extracting relevant information from unstructured medical documents contained in Electronic Health Records (EHR). However, most of the existing computer-based methods for clinical coding act as "black boxes", without giving a detailed description of the reasons for the clinical-coding assignments, which greatly limits their applicability to real-world medical scenarios. The objective of this study is to use transformer-based models to effectively tackle explainable clinical-coding. In this way, we require the models to perform the assignments of clinical codes to medical cases, but also to provide the reference in the text that justifies each coding assignment. METHODS: We examine the performance of 3 transformer-based architectures on 3 different explainable clinical-coding tasks. For each transformer, we compare the performance of the original general-domain version with an in-domain version of the model adapted to the specificities of the medical domain. We address the explainable clinical-coding problem as a dual medical named entity recognition (MER) and medical named entity normalization (MEN) task. For this purpose, we have developed two different approaches, namely a multi-task and a hierarchical-task strategy. RESULTS: For each analyzed transformer, the clinical-domain version significantly outperforms the corresponding general domain model across the 3 explainable clinical-coding tasks analyzed in this study. Furthermore, the hierarchical-task approach yields a significantly superior performance than the multi-task strategy. Specifically, the combination of the hierarchical-task strategy with an ensemble approach leveraging the predictive capabilities of the 3 distinct clinical-domain transformers, yields the best obtained results, with f1-score, precision and recall of 0.852, 0.847 and 0.849 on the Cantemist-Norm task and 0.718, 0.566 and 0.633 on the CodiEsp-X task, respectively. CONCLUSIONS: By separately addressing the MER and MEN tasks, as well as by following a context-aware text-classification approach to tackle the MEN task, the hierarchical-task approach effectively reduces the intrinsic complexity of explainable clinical-coding, leading the transformers to establish new SOTA performances for the predictive tasks considered in this study. In addition, the proposed methodology has the potential to be applied to other clinical tasks that require both the recognition and normalization of medical entities.