Assessing micro- vs macro-costing approaches for treating appendicitis in children with appendicectomy or non-operatively.

OBJECTIVES: We conducted a health economic sub-study within a feasibility RCT comparing a non-operative treatment pathway as an alternative to appendicectomy for the treatment of uncomplicated acute appendicitis in children. The objectives were to understand and assess data collection tools and methods and to determine indicative costs and benefits assessing the feasibility of conducting a full economic evaluation within the definitive trial. METHODS: We compared different methods of estimating treatment costs including micro-costing, hospital administrative data (PLICS) and health system (NHS) reference costs. We compared two different HRQoL instruments (CHU-9D and EQ-5D-5L) in terms of data completeness and sensitivity to change over time, including potential ceiling effects. We also explored how the timing of data collection and duration of the analysis could affect QALYs (Quality Adjusted Life Years) and the results of the cost-utility analysis (CUA) within the future RCT. RESULTS: Using a micro-costing approach, the total per treatment costs were in alignment with hospital administrative data (PLICS). Average health system reference cost data (macro-costing using NHS costs) could potentially underestimate these treatment costs, particularly for non-operative treatment. Costs incurred following hospital discharge in the primary care setting were minimal, and limited family borne costs were reported by parents/carers. While both HRQoL instruments performed relatively well, our results highlight the problem of ceiling effect and the importance of the timing of data collection and the duration of the analysis in any future assessment using QALYs and CUA. CONCLUSIONS: We highlighted the importance of obtaining accurate individual-patient cost data when conducting economic evaluations. Our results suggest that timing of data collection and duration of the assessment are important considerations when evaluating cost-effectiveness and reporting cost per QALY. CLINICAL TRIAL REGISTRATION: Current Controlled Trials ISRCTN15830435.

Benchmarking short-term postoperative mortality across neurosurgery units: is hospital administrative data good enough for risk-adjustment?

BACKGROUND: Surgical mortality indicators should be risk-adjusted when evaluating the performance of organisations. This study evaluated the performance of risk-adjustment models that used English hospital administrative data for 30-day mortality after neurosurgery. METHODS: This retrospective cohort study used Hospital Episode Statistics (HES) data from 1 April 2013 to 31 March 2018. Organisational-level 30-day mortality was calculated for selected subspecialties (neuro-oncology, neurovascular and trauma neurosurgery) and the overall cohort. Risk adjustment models were developed using multivariable logistic regression and incorporated various patient variables: age, sex, admission method, social deprivation, comorbidity and frailty indices. Performance was assessed in terms of discrimination and calibration. RESULTS: The cohort included 49,044 patients. Overall, 30-day mortality rate was 4.9%, with unadjusted organisational rates ranging from 3.2 to 9.3%. The variables in the best performing models varied for the subspecialties; for trauma neurosurgery, a model that included deprivation and frailty had the best calibration, while for neuro-oncology a model with these variables plus comorbidity performed best. For neurovascular surgery, a simple model of age, sex and admission method performed best. Levels of discrimination varied for the subspecialties (range: 0.583 for trauma and 0.740 for neurovascular). The models were generally well calibrated. Application of the models to the organisation figures produced an average (median) absolute change in mortality of 0.33% (interquartile range (IQR) 0.15-0.72) for the overall cohort model. Median changes for the subspecialty models were 0.29% (neuro-oncology, IQR 0.15-0.42), 0.40% (neurovascular, IQR 0.24-0.78) and 0.49% (trauma neurosurgery, IQR 0.23-1.68). CONCLUSIONS: Reasonable risk-adjustment models for 30-day mortality after neurosurgery procedures were possible using variables from HES, although the models for trauma neurosurgery performed less well. Including a measure of frailty often improved model performance.

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.

Shifts in Alcohol-Related Diagnoses After the Introduction of International Classification of Diseases, Tenth Revision, Clinical Modification Coding in U.S. Hospitals: Implications for Epidemiologic Research.

BACKGROUND: In October 2015, the United States transitioned healthcare diagnosis codes from the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM), to the Tenth Revision (ICD-10-CM). Trend analyses of alcohol-related stays could show discontinuities solely from the change in classification systems. This study examined the impact of the ICD-10-CM coding system on estimates of hospital stays involving alcohol-related diagnoses. METHODS: This analysis used 2014 to 2017 administrative data from the Agency for Healthcare Research and Quality Healthcare Cost and Utilization Project State Inpatient Databases for 17 states. Quarterly ICD-9-CM data from second quarter 2014 through third quarter 2015 were concatenated with ICD-10-CM data from fourth quarter 2015 through first quarter 2017. Quarterly counts of alcohol-related stays were examined overall and then by 6 diagnostic subgroups: withdrawal, abuse, dependence, alcohol-induced mental disorders (AIMD), nonpsychiatric alcohol-induced disease, and intoxication or toxic effects. Within each group, we calculated the difference in the average number of stays between ICD-9-CM and ICD-10-CM coding periods. RESULTS: On average, the number of stays involving any alcohol-related diagnosis in the 6 quarters before and after the ICD-10-CM transition was stable. However, substantial shifts in stays occurred for alcohol abuse, AIMD, and intoxication or toxic effects. For example, the average quarterly number of stays involving AIMD was 170.7% higher in the ICD-10-CM period than in the ICD-9-CM period. This increase was driven in large part by 1 ICD-10-CM code, Alcohol use, unspecified with unspecified alcohol-induced disorder. CONCLUSIONS: Researchers conducting trend analyses of inpatient stays involving alcohol-related diagnoses should consider how ongoing modifications in the ICD-10-CM code system and coding guidelines might affect their work. An advisable approach for trend analyses across the ICD-10-CM transition is to aggregate diagnosis codes into broader, clinically meaningful groups-including a single global group that encompasses all alcohol-related stays-and then to select diagnostic groupings that minimize discontinuities between the 2 coding systems while providing useful information on this important indicator of population health.

Development and validation of a coding framework to identify severe acute toxicity from systemic anti-cancer therapy using hospital administrative data.

BACKGROUND: The capture of toxicities from systemic anti-cancer therapy (SACT) in real-world data will complement results from clinical trials. The aim of this study was to develop and validate a comprehensive coding framework to identify severe acute toxicity in hospital administrative data. METHODS: A coding framework was developed to identify diagnostic codes representing severe acute toxicity in hospital administrative data. The coding framework was validated on a sample of 23,265 colon cancer patients treated in the English National Health Service between 1 June 2014 and 31 December 2017. This involved comparing individual toxicities according to the receipt of SACT and according to different SACT regimens, as well as assessing the associations of predictive factors and outcomes with toxicity. RESULTS: The severe acute toxicities captured by the developed coding framework were shown to vary across clinical groups with an overall rate of 26.4% in the adjuvant cohort, 53.4% in the metastatic cohort, and 12.5% in the comparison group receiving no chemotherapy. Results were in line with regimen-specific findings from clinical trials. For example, patients receiving additional bevacizumab had higher rates of bleeding (12.5% vs. 2.7%), gastrointestinal perforation (5.6% vs. 2.9%) and fistulation (1.4% vs. 0.5%), and allergic drug reactions (1.4% vs. 0.5%). Severe acute toxicity was associated with pre-existing renal (p = 0.001) and cardiac disease (p = 0.038), and urgency of surgery (p = 0.004). Severe toxicity also predicted lower rates of completion of chemotherapy (p = <0.001) and an increased likelihood of altered administration route (p = <0.001). CONCLUSION: These results demonstrate that the developed coding framework captures severe acute toxicities from hospital administrative data of colon cancer patients. A similar approach can be used for patients with other cancer types, receiving different regimens. Toxicity captured in administrative data can be used to compare treatment outcomes, inform clinical decision making, and provide opportunities for benchmarking and provider performance monitoring.

Bipolar disorder and medical comorbidities: A Portuguese population-based observational retrospective study (2008-2015).

BACKGROUND: This study aimed to use the Charlson Comorbidity Index (CCI) to assess the prevalence of medical comorbidities among hospitalization episodes with a primary Bipolar Disorder (BD) diagnosis, and to analyze its association with hospitalization outcomes. METHODS: A population-based observational retrospective study was conducted using a Portuguese administrative database containing all mainland public hospitalizations. From 2008-2015, hospitalization episodes with a primary diagnosis of BD were analysed. Outcomes included: length of stay (LoS), in-hospital mortality and discharge destination. RESULTS: Overall, 20807 hospitalization episodes were analysed. Mean±standard deviation age at admission was 47.9±14.3 years, and these episodes mostly refer to women's admissions (66.6%). Median (1st quartile; 3rd quartile) LoS was 16.0 (9.0; 25.0) days. A total of 2145 (10.3%) episodes had ≥1 CCI comorbidities registered, being diabetes the most prevalent. LoS was significantly higher for episodes with secondary diagnoses of congestive heart failure, cerebrovascular disease, dementia, diabetes, renal disease and malignancy (all p<0.05). Episodes with a registry of myocardial infarction, peripheral vascular disease, malignancy and renal disease diagnoses had higher in-hospital mortality. LIMITATIONS: Limitations include the use of data registered for administrative reasons rather than research purposes, and the analysis of hospitalization episodes, instead of patients. CONCLUSIONS: In this Portuguese nationwide study, greater comorbidity had a measurable impact on BD hospitalization outcomes. During the study period the prevalence of CCI comorbidities rose from 8.1% to 17.4%, which may reflect the overall increasing quality of hospital-coded data in Portugal throughout the years. The detection and timely management of medical comorbid conditions will likely prevent the high BD medical burden.

Seasonal variation of diabetes with hyperosmolarity hospitalizations and its characteristics in mainland Portugal.

AIMS: The family physician devotes a part of his care to the surveillance of diabetic patients. Hyperosmolarity is a severe acute complication. The aim of this study was to analyse seasonal variation of type 2 diabetes with hyperosmolarity hospitalizations, regarding their occurrence, mortality, length of stay, Charlson comorbidity index and its factors. METHODS: The authors analysed all hospitalizations in Portuguese Mainland public sector hospitals between 2000 and 2015 with primary diagnosis of type 2 diabetes with hyperosmolarity (ICD-9-CM codes 250.20 or 250.22), using a national administrative database. Cases were classified into four seasons according to date of admission. The authors compared the occurrence, length of stay, in-hospital mortality and Charlson comorbidity index and its factors. RESULTS: A total of 6596 hospitalization episodes were included. The authors found that admissions occurred more in winter, being 23% more common. No seasonal statistically significant differences were found considering the other variables. CONCLUSIONS: There is an increased occurrence of this acute metabolic complication during the winter in patients with type 2 diabetes. These results should be taken into account by the family physician when planning surveillance to this risk group.

Observational study to determine the utility of hospital administrative data to support case finding of English patients at higher risk of severe healthcare-related harm.

OBJECTIVES: To identify ways of using routine hospital data to improve the efficiency of retrospective reviews of case records for identifying avoidable severe harm DESIGN: Development and testing of thresholds and criteria for two indirect indicators of healthcare-related harm (long length of stay (LOS) and emergency readmission) to determine the yield of specified harms coded in Hospital Episode Statistics (HES). SETTING: Acute National Health Service hospitals in England. PARTICIPANTS: HES for acute myocardial infarction (AMI), bowel cancer surgery and hip replacement admissions from 2014 to 2015. INTERVENTIONS: Case-mix-adjusted linear regression models were used to determine expected LOS. Different thresholds were examined to determine the association with harm. Screening criteria for readmission included time to readmission, length of readmission and diagnoses in initial admission and readmission. The association with harm was examined for each criterion. RESULTS: The proportions of AMI cases with a harm code increased from 14% among all cases to 47% if a threshold of three times the expected LOS was used. For hip replacement the respective increase was from 10% to 51%. However as the number of patients at these higher thresholds was small, the overall proportion of harm identified is relatively small (15%, 19%, 9% and 8% among AMI, urgent bowel surgery, elective bowel surgery and hip replacement cohorts, respectively). Selection of the time to readmission had an effect on the yield of harms but this varied with condition. At least 50% of surgical patients had a harm code if readmitted within 7 days compared with 21% of patients with AMI. CONCLUSIONS: Our approach would select a substantial number of patients for case record review. Many of these cases would contain no evidence of healthcare-related harm. In practice, Trusts may choose how many reviews it is feasible to do in advance and then select random samples of cases that satisfy the screening criteria.

Using hospital administrative data to infer patient-patient contact via the consistent co-presence algorithm.

In health care settings, patients who are physically proximate to other patients (co-presence) for a meaningful amount of time may have differential health outcomes depending on who they are in contact with. How to best measure this co-presence, however is an open question and previous approaches have limitations that may make them inappropriate for complex health care settings. Here, we introduce a novel method which we term "consistent co-presence", that implicitly models the many complexities of patient scheduling and movement through a hospital by randomly perturbing the timing of patients' entry time into the health care system. This algorithm generates networks that can be employed in models of patient outcomes, such as 1-year mortality, and are preferred over previously established alternative algorithms from a model comparison perspective. These results indicate that consistent co-presence retains meaningful information about patient-patient interaction, which may affect outcomes relevant to health care practice. Furthermore, the generalizabiity of this approach allows it to be applied to a wide variety of complex systems.

Hospital admissions 2000-2014: A retrospective analysis of 288 096 events in patients with dementia.

INTRODUCTION: Dementia is a leading cause of disability worldwide. It is associated with an increased risk of hospitalization, imposing a significant burden on healthcare systems. The evidence on the long-term evolution of this issue and broadly on healthcare systems is currently limited. This study aims to describe the hospitalizations of people who received a diagnosis of dementia admitted to public general hospitals in a western European country with universal health coverage, over more than a decade. METHODS: This retrospective observational study analyzed all inpatient episodes from 2000 to 2014 with a primary or secondary diagnosis of dementia using a national hospitalization database from mainland Portuguese public hospitals. RESULTS: A total of 288 096 hospital admissions were registered. Hospitalization rates increased 4.7 times throughout the study period. Pneumonia and urinary tract infections were the most frequent main diagnoses, while dementia itself was the cause of admission in a minority (6.8%) of cases. Cerebrovascular disease, diabetes without chronic complications, and congestive heart failure were the most prevalent comorbidities; 5.9% of patients with dementia admitted to hospital underwent a surgical procedure, orthopedic surgeries being the most frequent. The median length of hospital stay was 8.0 days, and in-hospital mortality rate was 16.1%. CONCLUSIONS: Dementia patients represent a significant amount of hospital admissions. Most leading causes of hospital admissions are preventable if timely diagnosed and could be effectively managed in the outpatient setting. These findings may be useful for healthcare resource planning and allocation. Further research should drive evidence-based reorganization of health care systems.

Transfer Frequency as a Measure of Hospital Capability and Regionalization.

OBJECTIVE: To provide metrics for quantifying the capability of hospitals and the degree of care regionalization. DATA SOURCE: Administrative database covering more than 10 million hospital encounters during a 3-year period (2012-2014) in Massachusetts. PRINCIPAL FINDINGS: We calculated the condition-specific probabilities of transfer for all acute care hospitals in Massachusetts and devised two new metrics, the Hospital Capability Index (HCI) and the Regionalization Index (RI), for analyzing hospital systems. The HCI had face validity, accurately differentiating academic, teaching, and community hospitals of varying size. Individual hospital capabilities were clearly revealed in "fingerprints" of their condition-specific transfer behavior. The RI also performed well, with those of specific conditions successfully quantifying the concentration of care arising from regulatory and public health activity. The median RI of all conditions within the Massachusetts health care system was 0.21 (IQR, 0.13-0.36), with a long tail of conditions that were very highly regionalized. Application of the HCI and RI metrics together across the entire state identified the degree of interdependence among its hospitals. CONCLUSIONS: Condition-specific transfer activity, as captured in the HCI and RI, provides quantitative measures of hospital capability and regionalization of care.

Enhancing Clinical Content and Race/Ethnicity Data in Statewide Hospital Administrative Databases: Obstacles Encountered, Strategies Adopted, and Lessons Learned.

OBJECTIVES: Eight grant teams used Agency for Healthcare Research and Quality infrastructure development research grants to enhance the clinical content of and improve race/ethnicity identifiers in statewide all-payer hospital administrative databases. PRINCIPAL FINDINGS: Grantees faced common challenges, including recruiting data partners and ensuring their continued effective participation, acquiring and validating the accuracy and utility of new data elements, and linking data from multiple sources to create internally consistent enhanced administrative databases. Successful strategies to overcome these challenges included aggressively engaging with providers of critical sources of data, emphasizing potential benefits to participants, revising requirements to lessen burdens associated with participation, maintaining continuous communication with participants, being flexible when responding to participants' difficulties in meeting program requirements, and paying scrupulous attention to preparing data specifications and creating and implementing protocols for data auditing, validation, cleaning, editing, and linking. In addition to common challenges, grantees also had to contend with unique challenges from local environmental factors that shaped the strategies they adopted. CONCLUSIONS: The creation of enhanced administrative databases to support comparative effectiveness research is difficult, particularly in the face of numerous challenges with recruiting data partners such as competing demands on information technology resources. Excellent communication, flexibility, and attention to detail are essential ingredients in accomplishing this task. Additional research is needed to develop strategies for maintaining these databases when initial funding is exhausted.

Influence of psychiatric disorders on surgical outcomes and care resource use in Japan.

OBJECTIVE: The aim of this study was to quantify the effects of psychiatric disorders on major surgery outcomes and care resource use. METHODS: This study adopted a retrospective cohort study design. The samples consisted of hospital stays. Subjects were patients who had undergone major surgery. We used multilevel regression analysis to quantify the influence of psychiatric disorders on major surgery outcomes and care resource use. RESULTS: The total number of hospital stays included in the study was 5569, of which 250 were patients with psychiatric disorders. Compared with those without psychiatric disorders, those with schizophrenia had a significantly higher risk of complications, and those with neurotic disorder tended to have fewer complications. Total cost was significantly higher for those with schizophrenia and mood disorder and significantly lower in those with neurotic disorder. Lengths of stay were significantly longer for those with schizophrenia and mood disorder but not for those with neurotic disorder. Post-surgical mortality was equivalent among those with any psychiatric disorder and among those without a psychiatric disorder. CONCLUSION: The study revealed that surgical outcomes and care resource use are differentiated by psychiatric disorders.

Coding for quality measurement: the relationship between hospital structural characteristics and coding accuracy from the perspective of quality measurement.

This study examines the relationship between hospital structural characteristics and coding accuracy from the perspective of quality measurement. To measure coding accuracy for quality measurement, the study utilizes the "present on admission" indicator, a data element in the New York state hospital administrative database. This data element is used by hospitals across New York state to indicate if a particular secondary diagnosis is "present on admission," "not present on admission," or "uncertain." Since the accurate distinction between comorbidities (present at admission) and complications (not present at admission,) is critical for risk adjustment in comparative hospital quality reports, this study uses the occurrence of the value "uncertain" in the "present on admission" indicator as the primary measure of coding accuracy. A lower occurrence of the value "uncertain" is considered to be reflective of better coding accuracy. Moreover, since coding accuracy of the "present on admission" indicator links back to the accuracy of physician documentation, a focus on the occurrence of the value "uncertain," also helps gain insight into physician documentation efficacy within the facility. By utilizing this approach, therefore, the study serves the twin purpose of 1) addressing the gap in the literature with respect to large-scale studies of "coding for quality," and 2) providing insight into the structural characteristics of institutions that are likely facing organizational challenges of physician documentation from the perspective of quality measurement.