Interpreting and coding causal relationships for quality and safety using ICD-11.

Many circumstances necessitate judgments regarding causation in health information systems, but these can be tricky in medicine and epidemiology. In this article, we reflect on what the ICD-11 Reference Guide provides on coding for causation and judging when relationships between clinical concepts are causal. Based on the use of different types of codes and the development of a new mechanism for coding potential causal relationships, the ICD-11 provides an in-depth transformation of coding expectations as compared to ICD-10. An essential part of the causal relationship interpretation relies on the presence of "connecting terms," key elements in assessing the level of certainty regarding a potential relationship and how to proceed in coding a causal relationship using the new ICD-11 coding convention of postcoordination (i.e., clustering of codes). In addition, determining causation involves using documentation from healthcare providers, which is the foundation for coding health information. The coding guidelines and examples (taken from the quality and patient safety domain) presented in this article underline how new ICD-11 features and coding rules will enhance future health information systems and healthcare.

Clinical and Health System Determinants of Venous Thromboembolism Event Rates After Hip Arthroplasty: An International Comparison.

BACKGROUND: Routinely collected hospital data provide increasing opportunities to assess the performance of health care systems. Several factors may, however, influence performance measures and their interpretation between countries. OBJECTIVE: We compared the occurrence of in-hospital venous thromboembolism (VTE) in patients undergoing hip replacement across 5 countries and explored factors that could explain differences across these countries. METHODS: We performed cross-sectional studies independently in 5 countries: Canada; France; New Zealand; the state of California; and Switzerland. We first calculated the proportion of hospital inpatients with at least one deep vein thrombosis (DVT) or pulmonary embolism by using numerator codes from the corresponding Patient Safety Indicator. We then compared estimates from each country against a reference value (benchmark) that displayed the baseline risk of VTE in such patients. Finally, we explored length of stay, number of secondary diagnoses coded, and systematic use of ultrasound to detect DVT as potential factors that could explain between-country differences. RESULTS: The rates of VTE were 0.16% in Canada, 1.41% in France, 0.84% in New Zealand, 0.66% in California, and 0.37% in Switzerland, while the benchmark was 0.58% (95% confidence interval, 0.35-0.81). Factors that could partially explain differences in VTE rates between countries were hospital length of stay, number of secondary diagnoses coded, and proportion of patients who received lower limb ultrasound to screen for DVT systematically before hospital discharge. An exploration of the French data showed that the systematic use of ultrasound may be associated with over detection of DVT but not pulmonary embolism. CONCLUSIONS: In-hospital VTE rates after arthroplasty vary widely across countries, and a combination of clinical, data-related, and health system factors explain some of the variations in VTE rates across countries.

A World Health Organization field trial assessing a proposed ICD-11 framework for classifying patient safety events.

OBJECTIVE: To assess the utility of the proposed World Health Organization (WHO)'s International Classification of Disease (ICD) framework for classifying patient safety events. SETTING: Independent classification of 45 clinical vignettes using a web-based platform. STUDY PARTICIPANTS: The WHO's multi-disciplinary Quality and Safety Topic Advisory Group. MAIN OUTCOME MEASURE(S): The framework consists of three concepts: harm, cause and mode. We defined a concept as 'classifiable' if more than half of the raters could assign an ICD-11 code for the case. We evaluated reasons why cases were nonclassifiable using a qualitative approach. RESULTS: Harm was classifiable in 31 of 45 cases (69%). Of these, only 20 could be classified according to cause and mode. Classifiable cases were those in which a clear cause and effect relationship existed (e.g. medication administration error). Nonclassifiable cases were those without clear causal attribution (e.g. pressure ulcer). Of the 14 cases in which harm was not evident (31%), only 5 could be classified according to cause and mode and represented potential adverse events. Overall, nine cases (20%) were nonclassifiable using the three-part patient safety framework and contained significant ambiguity in the relationship between healthcare outcome and putative cause. CONCLUSIONS: The proposed framework enabled classification of the majority of patient safety events. Cases in which potentially harmful events did not cause harm were not classifiable; additional code categories within the ICD-11 are one proposal to address this concern. Cases with ambiguity in cause and effect relationship between healthcare processes and outcomes remain difficult to classify.

Matching Registered Nurse services with changing care demands (MatchRN ): study protocol of a natural experiment multi-centre study.

AIM: The aim of this study was to examine how patient safety indicators and processes and structures of nursing care have changed since the 2012 introduction of Swiss Diagnosis-Related Groups. BACKGROUND: Diagnosis-Related Groups have been implemented worldwide; yet, research findings regarding their impact on efficiency and quality of care remain inconsistent. The Matching Registered Nurse Services with Changing Care Demands study will assess how structures, processes and patient and nurse outcomes have changed in Swiss acute care hospitals since the introduction of Swiss Diagnosis-Related Groups. DESIGN: A multi-centre observational study nested in a natural experiment. METHODS: To explore the effect of implementing Diagnosis-Related Groups in Switzerland we will compare nurse and patient survey data from 2010 with data from 2015 and eventually from 2017. Initially, we will match survey data from 78 medical and surgical units of 21 hospitals that participated in 2010 and 2015. Study variables related to structures and processes of nursing care (e.g. staffing/skill mix level, nurse work environment, rationing of nursing care), as well as patient and nurse outcomes, were assessed with well-established instruments. In 2017, a follow-up survey will be conducted to explore long-term implications. Furthermore, 6 years' medical and surgical patient discharge data (collected 2010-2015) will be analysed to assess changes in the severity of patient illness, length of stay and selected patient safety indicators. DISCUSSION: This study's results will provide evidence regarding Diagnosis-Related Groups influences on Swiss nursing services and patient safety outcomes.

[The vision of nursing provided by Léonie Chaptal: a strategic mistake in the development of nursing science in France ?] / La vision des soins portée par Léonie Chaptal : une erreur stratégique dans le développement de la discipline des sciences infirmières en France ?

The issue of the development of the discipline displayed by nursing sciences in France is crucial, especially since the Law of Modernization of the Health System has established a legal framework for advanced practice since 2016. This article presents a discussion on the important role that Léonie Chaptal has played in the development of nursing care in France, based on a purely professional vision, guaranteeing the subordination of the nurses to the medical physicians the in the spirit of the law of November 30, 1892 which had established a stranglehold of medical profession on health. We have drawn from this discussion some lessons that we consider essential for the future development of a true discipline of nursing in France, in the context of the deep organizational transformation initiated by the health system.

[Improve the extravasation assessment of peripheral venous catheter by nurses, with the establishment of a standardized instrument and suitable for children] / Amélioration de la mesure d'extravasation des cathéters veineux périphériques par les infirmières en pédiatrie, grâce à l'utilisation d'un instrument d'évaluation spécifique.

In pediatric and neonatal intensive care units, severity assessment of extravasation is difficult, considering the specificity of this population. The purpose of this study was to demonstrate the ability to improve the measurement of extravasation by nurses with the establishment of a standardized instrument and suitable for children. 66 nurses, randomly assigned to two groups, assessed the severity of extravasations using 15 clinical vignettes.The intervention group with the Pediatric Peripheral Intravenous Infiltration Scale (PIV Scale) (n=33) and the control group based on clinical judgment only (n=33). The reference were obtained from a group of experts. For both groups, concordance and sensitivity were calculated. Concordance and sensitivity were improved by the use of the PIV scale κ=0,62 (IC 95% ; 0,57-0,67) vs κ=0,51 (IC 95 % ; 0,45-0,57), and (69 %) vs (60 %) (p<0,001), respectively. Severity assessment of extravasation on peripheral venous accesses by nurses was improved with the use of the PIV scale, compared to clinical judgment. As this study was based on clinical vignettes, further studies are needed to confirm these results in clinical setting.

Digital Health Education for the Future: The SaNuRN (Santé Numérique Rouen-Nice) Consortium's Journey.

SaNuRN is a five-year project by the University of Rouen Normandy (URN) and the Côte d'Azur University (CAU) consortium to optimize digital health education for medical and paramedical students, professionals, and administrators. The project includes a skills framework, training modules, and teaching resources. In 2027, SaNuRN is expected to train a significant portion of the 400,000 health and paramedical professions students at the French national level. Our purpose is to give a synopsis of the SaNuRN initiative, emphasizing its novel educational methods and how they will enhance the delivery of digital health education. Our goals include showcasing SaNuRN as a comprehensive program consisting of a proficiency framework, instructional modules, and educational materials and explaining how SaNuRN is implemented in the participating academic institutions. SaNuRN is a project aimed at educating and training health-related and paramedics students in digital health. The project results from a cooperative effort between URN and CAU, covering four French departments. The project is based on the French National Referential on Digital Health (FNRDH), which defines the skills and competencies to be acquired and validated by every student in the health, paramedical, and social professions curricula. The SaNuRN team is currently adapting the existing URN and CAU syllabi to FNRDH and developing short-duration video capsules of 20 to 30 minutes to teach all the relevant material. The project aims to ensure that the largest student population earns the necessary skills, and it has developed a two-tier system involving facilitators who will enable the efficient expansion of the project's educational outreach and support the students in learning the needed material efficiently. With a focus on real-world scenarios and innovative teaching activities integrating telemedicine devices and virtual professionals, SaNuRN is committed to enabling continuous learning for healthcare professionals in clinical practice. The SaNuRN team introduced new ways of evaluating healthcare professionals by shifting from a knowledge-based to a competencies-based evaluation, aligning with the Miller teaching pyramid and using the Objective Structured Clinical Examination and Script Concordance Test in digital health education. Drawing on the expertise of URN, CAU, and their public health and digital research laboratories and partners, the SaNuRN project represents a platform for continuous innovation, including telemedicine training and living labs with virtual and interactive professional activities. The SaNuRN project provides a comprehensive, personalized 30-hour training package for health and paramedical students, addressing all 70 FNRDH competencies. The program is enhanced using AI and NLP to create virtual patients and professionals for digital healthcare simulation. SaNuRN teaching materials are open-access. The project collaborates with academic institutions worldwide to develop educational material in digital health in English and multilingual formats. SaNuRN offers a practical and persuasive training approach to meet the current digital health education requirements.

Systematic scoping review of automated systems for the surveillance of healthcare-associated bloodstream infections related to intravascular catheters.

INTRODUCTION: Intravascular catheters are crucial devices in medical practice that increase the risk of healthcare-associated infections (HAIs), and related health-economic adverse outcomes. This scoping review aims to provide a comprehensive overview of published automated algorithms for surveillance of catheter-related bloodstream infections (CRBSI) and central line-associated bloodstream infections (CLABSI). METHODS: We performed a scoping review based on a systematic search of the literature in PubMed and EMBASE from 1 January 2000 to 31 December 2021. Studies were included if they evaluated predictive performance of automated surveillance algorithms for CLABSI/CRBSI detection and used manually collected surveillance data as reference. We assessed the design of the automated systems, including the definitions used to develop algorithms (CLABSI versus CRBSI), the datasets and denominators used, and the algorithms evaluated in each of the studies. RESULTS: We screened 586 studies based on title and abstract, and 99 were assessed based on full text. Nine studies were included in the scoping review. Most studies were monocentric (n = 5), and they identified CLABSI (n = 7) as an outcome. The majority of the studies used administrative and microbiological data (n = 9) and five studies included the presence of a vascular central line in their automated system. Six studies explained the denominator they selected, five of which chose central line-days. The most common rules and steps used in the algorithms were categorized as hospital-acquired rules, infection rules (infection versus contamination), deduplication, episode grouping, secondary BSI rules (secondary versus primary BSI), and catheter-associated rules. CONCLUSION: The automated surveillance systems that we identified were heterogeneous in terms of definitions, datasets and denominators used, with a combination of rules in each algorithm. Further guidelines and studies are needed to develop and implement algorithms to detect CLABSI/CRBSI, with standardized definitions, appropriate data sources and suitable denominators.

Cohort profile for development of machine learning models to predict healthcare-related adverse events (Demeter): clinical objectives, data requirements for modelling and overview of data set for 2016-2018.

PURPOSE: In-hospital health-related adverse events (HAEs) are a major concern for hospitals worldwide. In high-income countries, approximately 1 in 10 patients experience HAEs associated with their hospital stay. Estimating the risk of an HAE at the individual patient level as accurately as possible is one of the first steps towards improving patient outcomes. Risk assessment can enable healthcare providers to target resources to patients in greatest need through adaptations in processes and procedures. Electronic health data facilitates the application of machine-learning methods for risk analysis. We aim, first to reveal correlations between HAE occurrence and patients' characteristics and/or the procedures they undergo during their hospitalisation, and second, to build models that allow the early identification of patients at an elevated risk of HAE. PARTICIPANTS: 143 865 adult patients hospitalised at Grenoble Alpes University Hospital (France) between 1 January 2016 and 31 December 2018. FINDINGS TO DATE: In this set-up phase of the project, we describe the preconditions for big data analysis using machine-learning methods. We present an overview of the retrospective de-identified multisource data for a 2-year period extracted from the hospital's Clinical Data Warehouse, along with social determinants of health data from the National Institute of Statistics and Economic Studies, to be used in machine learning (artificial intelligence) training and validation. No supplementary information or evaluation on the part of medical staff will be required by the information system for risk assessment. FUTURE PLANS: We are using this data set to develop predictive models for several general HAEs including secondary intensive care admission, prolonged hospital stay, 7-day and 30-day re-hospitalisation, nosocomial bacterial infection, hospital-acquired venous thromboembolism, and in-hospital mortality.

Predictive performance of automated surveillance algorithms for intravascular catheter bloodstream infections: a systematic review and meta-analysis.

BACKGROUND: Intravascular catheter infections are associated with adverse clinical outcomes. However, a significant proportion of these infections are preventable. Evaluations of the performance of automated surveillance systems for adequate monitoring of central-line associated bloodstream infection (CLABSI) or catheter-related bloodstream infection (CRBSI) are limited. OBJECTIVES: We evaluated the predictive performance of automated algorithms for CLABSI/CRBSI detection, and investigated which parameters included in automated algorithms provide the greatest accuracy for CLABSI/CRBSI detection. METHODS: We performed a meta-analysis based on a systematic search of published studies in PubMed and EMBASE from 1 January 2000 to 31 December 2021. We included studies that evaluated predictive performance of automated surveillance algorithms for CLABSI/CRBSI detection and used manually collected surveillance data as reference. We estimated the pooled sensitivity and specificity of algorithms for accuracy and performed a univariable meta-regression of the different parameters used across algorithms. RESULTS: The search identified five full text studies and 32 different algorithms or study populations were included in the meta-analysis. All studies analysed central venous catheters and identified CLABSI or CRBSI as an outcome. Pooled sensitivity and specificity of automated surveillance algorithm were 0.88 [95%CI 0.84-0.91] and 0.86 [95%CI 0.79-0.92] with significant heterogeneity (I2 = 91.9, p < 0.001 and I2 = 99.2, p < 0.001, respectively). In meta-regression, algorithms that include results of microbiological cultures from specific specimens (respiratory, urine and wound) to exclude non-CRBSI had higher specificity estimates (0.92, 95%CI 0.88-0.96) than algorithms that include results of microbiological cultures from any other body sites (0.88, 95% CI 0.81-0.95). The addition of clinical signs as a predictor did not improve performance of these algorithms with similar specificity estimates (0.92, 95%CI 0.88-0.96). CONCLUSIONS: Performance of automated algorithms for detection of intravascular catheter infections in comparison to manual surveillance seems encouraging. The development of automated algorithms should consider the inclusion of results of microbiological cultures from specific specimens to exclude non-CRBSI, while the inclusion of clinical data may not have an added-value. Trail Registration Prospectively registered with International prospective register of systematic reviews (PROSPERO ID CRD42022299641; January 21, 2022). https://www.crd.york.ac.uk/prospero/display_record.php?ID=CRD42022299641.

Role of comorbidities in outcome prediction after status epilepticus.

Status epilepticus (SE) is associated with significant mortality and morbidity. A reliable prognosis may help better manage medical resources and treatment strategies. We examined the role of preexisting comorbidities on the outcome of patients with SE, an aspect that has received little attention to date. We prospectively studied incident SE episodes in 280 adults occurring over 55 months in our tertiary care hospital, excluding patients with postanoxic encephalopathy. Different models predicting mortality and return to clinical baseline at hospital discharge were compared, which included demographics, SE etiology, a validated clinical Status Epilepticus Severity Score (STESS), and comorbidities (assessed with the Charlson Comorbidity Index) as independent variables. The overall short-term mortality was 14%, and only half of patients returned to their clinical baseline. On bivariate analyses, age, STESS, potentially fatal etiologies, and number of preexisting comorbidities were all significant predictors of both mortality and return to clinical baseline. As compared with the simplest predictive model (including demographics and deadly etiology), adding SE severity and comorbidities resulted in an improved predictive performance (C statistics 0.84 vs. 0.77 for mortality, and 0.86 vs. 0.82. for return to clinical baseline); comorbidities, however, were not independently related to outcome. Considering comorbidities and clinical presentation, in addition to age and etiology, slightly improves the prediction of SE outcome with respect to both survival and functional status. This analysis also emphasizes the robust predictive role of etiology and age.

Symptomatic in-hospital deep vein thrombosis and pulmonary embolism following hip and knee arthroplasty among patients receiving recommended prophylaxis: a systematic review.

CONTEXT: Symptomatic venous thromboembolism (VTE) after total or partial knee arthroplasty (TPKA) and after total or partial hip arthroplasty (TPHA) are proposed patient safety indicators, but its incidence prior to discharge is not defined. OBJECTIVE: To establish a literature-based estimate of symptomatic VTE event rates prior to hospital discharge in patients undergoing TPHA or TPKA. DATA SOURCES: Search of MEDLINE, EMBASE, and the Cochrane Library (1996 to 2011), supplemented by relevant articles. STUDY SELECTION: Reports of incidence of symptomatic postoperative pulmonary embolism or deep vein thrombosis (DVT) before hospital discharge in patients who received VTE prophylaxis with either a low-molecular-weight heparin or a subcutaneous factor Xa inhibitor or oral direct inhibitor of factors Xa or IIa. DATA EXTRACTION AND SYNTHESIS: Meta-analysis of randomized clinical trials and observational studies that reported rates of postoperative symptomatic VTE in patients who received recommended VTE prophylaxis after undergoing TPHA or TPKA. Data were independently extracted by 2 analysts, and pooled incidence rates of VTE, DVT, and pulmonary embolism were estimated using random-effects models. RESULTS: The analysis included 44,844 cases provided by 47 studies. The pooled rates of symptomatic postoperative VTE before hospital discharge were 1.09% (95% CI, 0.85%-1.33%) for patients undergoing TPKA and 0.53% (95% CI, 0.35%-0.70%) for those undergoing TPHA. The pooled rates of symptomatic DVT were 0.63% (95% CI, 0.47%-0.78%) for knee arthroplasty and 0.26% (95% CI, 0.14%-0.37%) for hip arthroplasty. The pooled rates for pulmonary embolism were 0.27% (95% CI, 0.16%-0.38%) for knee arthroplasty and 0.14% (95% CI, 0.07%-0.21%) for hip arthroplasty. There was significant heterogeneity for the pooled incidence rates of symptomatic postoperative VTE in TPKA studies but less heterogeneity for DVT and pulmonary embolism in TPKA studies and for VTE, DVT, and pulmonary embolism in TPHA studies. CONCLUSION: Using current VTE prophylaxis, approximately 1 in 100 patients undergoing TPKA and approximately 1 in 200 patients undergoing TPHA develops symptomatic VTE prior to hospital discharge.

Updating and validating the Charlson comorbidity index and score for risk adjustment in hospital discharge abstracts using data from 6 countries.

With advances in the effectiveness of treatment and disease management, the contribution of chronic comorbid diseases (comorbidities) found within the Charlson comorbidity index to mortality is likely to have changed since development of the index in 1984. The authors reevaluated the Charlson index and reassigned weights to each condition by identifying and following patients to observe mortality within 1 year after hospital discharge. They applied the updated index and weights to hospital discharge data from 6 countries and tested for their ability to predict in-hospital mortality. Compared with the original Charlson weights, weights generated from the Calgary, Alberta, Canada, data (2004) were 0 for 5 comorbidities, decreased for 3 comorbidities, increased for 4 comorbidities, and did not change for 5 comorbidities. The C statistics for discriminating in-hospital mortality between the new score generated from the 12 comorbidities and the Charlson score were 0.825 (new) and 0.808 (old), respectively, in Australian data (2008), 0.828 and 0.825 in Canadian data (2008), 0.878 and 0.882 in French data (2004), 0.727 and 0.723 in Japanese data (2008), 0.831 and 0.836 in New Zealand data (2008), and 0.869 and 0.876 in Swiss data (2008). The updated index of 12 comorbidities showed good-to-excellent discrimination in predicting in-hospital mortality in data from 6 countries and may be more appropriate for use with more recent administrative data.

Second-line status epilepticus treatment: comparison of phenytoin, valproate, and levetiracetam.

PURPOSE: Phenytoin (PHT), valproic acid (VPA), or levetiracetam (LEV) are commonly used as second-line treatment of status epilepticus (SE), but comparative studies are not available. METHODS: Among 279 adult SE episodes identified prospectively in our tertiary care hospital over 4 years, we retrospectively identified 187 episodes in which PHT, VPA, or LEV were given after benzodiazepines. Patients with postanoxic SE were not included. Demographics, clinical SE features, failure of second-line treatment to control SE, new handicap, and mortality at hospital discharge were assessed. Uni- and multivariable statistical analyses were applied to compare the three agents. KEY FINDINGS: Each compound was used in about one third of SE episodes. VPA failed to control SE in 25.4%, PHT in 41.4%, and LEV in 48.3% of episodes in which these were prescribed. A deadly etiology was more frequent in the VPA group, whereas SE episodes tended to be more severe in the PHT group. After adjustment for these known SE outcome predictors, LEV failed more often than VPA [odds ratio (OR) 2.69; 95% confidence interval (CI) 1.19-6.08]; 16.8% (95% CI: 6.0-31.4%) of second-line treatment failures could be attributed to LEV. PHT was not statistically different from the other two compounds. Second-line treatment did not seem to influence new handicap and mortality, whereas etiology and the SE Severity Score (STESS) were robust independent predictors. SIGNIFICANCE: Even without significant differences on outcome at discharge, LEV seems less efficient than VPA to control SE after benzodiazepines. A prospective comparative trial is needed to address this potentially concerning finding.

Early-onset ventilator-associated pneumonia incidence in intensive care units: a surveillance-based study.

BACKGROUND: The incidence of ventilator-associated pneumonia (VAP) within the first 48 hours of intensive care unit (ICU) stay has been poorly investigated. The objective was to estimate early-onset VAP occurrence in ICUs within 48 hours after admission. METHODS: We analyzed data from prospective surveillance between 01/01/2001 and 31/12/2009 in 11 ICUs of Lyon hospitals (France). The inclusion criteria were: first ICU admission, not hospitalized before admission, invasive mechanical ventilation during first ICU day, free of antibiotics at admission, and ICU stay ≥ 48 hours. VAP was defined according to a national protocol. Its incidence was the number of events per 1,000 invasive mechanical ventilation-days. The Poisson regression model was fitted from day 2 (D2) to D8 to incident VAP to estimate the expected VAP incidence from D0 to D1 of ICU stay. RESULTS: Totally, 367 (10.8%) of 3,387 patients in 45,760 patient-days developed VAP within the first 9 days. The predicted cumulative VAP incidence at D0 and D1 was 5.3 (2.6-9.8) and 8.3 (6.1-11.1), respectively. The predicted cumulative VAP incidence was 23.0 (20.8-25.3) at D8. The proportion of missed VAP within 48 hours from admission was 11% (9%-17%). CONCLUSIONS: Our study indicates underestimation of early-onset VAP incidence in ICUs, if only VAP occurring ≥ 48 hours are considered to be hospital-acquired. Clinicians should be encouraged to develop a strategy for early detection after ICU admission.

Improved accuracy of co-morbidity coding over time after the introduction of ICD-10 administrative data.

BACKGROUND: Co-morbidity information derived from administrative data needs to be validated to allow its regular use. We assessed evolution in the accuracy of coding for Charlson and Elixhauser co-morbidities at three time points over a 5-year period, following the introduction of the International Classification of Diseases, 10th Revision (ICD-10), coding of hospital discharges. METHODS: Cross-sectional time trend evaluation study of coding accuracy using hospital chart data of 3'499 randomly selected patients who were discharged in 1999, 2001 and 2003, from two teaching and one non-teaching hospital in Switzerland. We measured sensitivity, positive predictive and Kappa values for agreement between administrative data coded with ICD-10 and chart data as the 'reference standard' for recording 36 co-morbidities. RESULTS: For the 17 the Charlson co-morbidities, the sensitivity - median (min-max) - was 36.5% (17.4-64.1) in 1999, 42.5% (22.2-64.6) in 2001 and 42.8% (8.4-75.6) in 2003. For the 29 Elixhauser co-morbidities, the sensitivity was 34.2% (1.9-64.1) in 1999, 38.6% (10.5-66.5) in 2001 and 41.6% (5.1-76.5) in 2003. Between 1999 and 2003, sensitivity estimates increased for 30 co-morbidities and decreased for 6 co-morbidities. The increase in sensitivities was statistically significant for six conditions and the decrease significant for one. Kappa values were increased for 29 co-morbidities and decreased for seven. CONCLUSIONS: Accuracy of administrative data in recording clinical conditions improved slightly between 1999 and 2003. These findings are of relevance to all jurisdictions introducing new coding systems, because they demonstrate a phenomenon of improved administrative data accuracy that may relate to a coding 'learning curve' with the new coding system.