Directory of clinical databases: improving and promoting their use

Table 0; documenting the steps to go from clinical database to research dataset

OBJECTIVES

Data-driven decision support tools have been increasingly recognized to transform health care. However, such tools are often developed on predefined research datasets without adequate knowledge of the origin of this data and how it was selected. How a dataset is extracted from a clinical database can profoundly impact the validity, interpretability and interoperability of the dataset, and downstream analyses, yet is rarely reported. Therefore, we present a case study illustrating how a definitive patient list was extracted from a clinical source database and how this can be reported.

STUDY DESIGN AND SETTING

A single-center observational study was performed at an academic hospital in the Netherlands to illustrate the impact of selecting a definitive patient list for research from a clinical source database, and the importance of documenting this process. All admissions from the critical care database admitted between January 1, 2013, and January 1, 2023, were used.

RESULTS

An interdisciplinary team collaborated to identify and address potential sources of data insufficiency and uncertainty. We demonstrate a stepwise data preparation process, reducing the clinical source database of 54,218 admissions to a definitive patient list of 21,553 admissions. Transparent documentation of the data preparation process improves the quality of the definitive patient list before analysis of the corresponding patient data. This study generated seven important recommendations for preparing observational health-care data for research purposes.

CONCLUSION

Documenting data preparation is essential for understanding a research dataset originating from a clinical source database before analyzing health-care data. The findings contribute to establishing data standards and offer insights into the complexities of preparing health-care data for scientific investigation. Meticulous data preparation and documentation thereof will improve research validity and advance critical care.

National ICU Registries as Enablers of Clinical Research and Quality Improvement

OBJECTIVES

Clinical quality registries (CQRs) have been implemented worldwide by several medical specialties aiming to generate a better characterization of epidemiology, treatments, and outcomes of patients. National ICU registries were created almost 3 decades ago to improve the understanding of case-mix, resource use, and outcomes of critically ill patients. This narrative review describes the challenges, proposed solutions, and evidence generated by National ICU registries as facilitators for research and quality improvement.

DATA SOURCES

English language articles were identified in PubMed using phrases related to ICU registries, CQRs, outcomes, and case-mix.

STUDY SELECTION

Original research, review articles, letters, and commentaries, were considered.

DATA EXTRACTION

Data from relevant literature were identified, reviewed, and integrated into a concise narrative review.

DATA SYNTHESIS

CQRs have been implemented worldwide by several medical specialties aiming to generate a better characterization of epidemiology, treatments, and outcomes of patients. National ICU registries were created almost 3 decades ago to improve the understanding of case-mix, resource use, and outcomes of critically ill patients. The initial experience in European countries and in Oceania ensured that through locally generated data, ICUs could assess their performances by using risk-adjusted measures and compare their results through fair and validated benchmarking metrics with other ICUs contributing to the CQR. The accomplishment of these initiatives, coupled with the increasing adoption of information technology, resulted in a broad geographic expansion of CQRs as well as their use in quality improvement studies, clinical trials as well as international comparisons, and benchmarking for ICUs.

CONCLUSIONS

ICU registries have provided increased knowledge of case-mix and outcomes of ICU patients based on real-world data and contributed to improve care delivery through quality improvement initiatives and trials. Recent increases in adoption of new technologies (i.e., cloud-based structures, artificial intelligence, machine learning) will ensure a broader and better use of data for epidemiology, healthcare policies, quality improvement, and clinical trials.

Collaboration for Research, Implementation and Training in Critical Care - Asia Investigators, Pisani L, Rashan T, Shamal M, Ghose A, Kumar Tirupakuzhi Vijayaraghavan B, Tripathy S, Aryal D, Hashmi M, Nor B, Lam Minh Y, Dondorp AM, Haniffa R, Beane A. Performance evaluation of a multinational data platform for critical care in Asia

Background: The value of medical registries strongly depends on the quality of the data collected. This must be objectively measured before large clinical databases can be promoted for observational research, quality improvement, and clinical trials. We aimed to evaluate the quality of a multinational intensive care unit (ICU) network of registries of critically ill patients established in seven Asian low- and middle-income countries (LMICs). Methods: The Critical Care Asia federated registry platform enables ICUs to collect clinical, outcome and process data for aggregate and unit-level analysis. The evaluation used the standardised criteria of the Directory of Clinical Databases (DoCDat) and a framework for data quality assurance in medical registries. Six reviewers assessed structure, coverage, reliability and validity of the ICU registry data. Case mix and process measures on patient episodes from June to December 2020 were analysed. Results: Data on 20,507 consecutive patient episodes from 97 ICUs in Afghanistan, Bangladesh, India, Malaysia, Nepal, Pakistan and Vietnam were included. The quality level achieved according to the ten prespecified DoCDat criteria was high (average score 3.4 out of 4) as was the structural and organizational performance -- comparable to ICU registries in high-income countries. Identified strengths were types of variables included, reliability of coding, data completeness and validation. Potential improvements included extension of national coverage, optimization of recruitment completeness validation in all centers and the use of interobserver reliability checks. Conclusions: The Critical Care Asia platform evaluates well using standardised frameworks for data quality and equally to registries in resource-rich settings.

Collaboration for Research, Implementation and Training in Critical Care - Asia Investigators, Pisani L, Rashan T, Shamal M, Ghose A, Kumar Tirupakuzhi Vijayaraghavan B, Tripathy S, Aryal D, Hashmi M, Nor B, Lam Minh Y, Dondorp AM, Haniffa R, Beane A. Performance evaluation of a multinational data platform for critical care in Asia

Background: The value of medical registries strongly depends on the quality of the data collected. This must be objectively measured before large clinical databases can be promoted for observational research, quality improvement, and clinical trials. We aimed to evaluate the quality of a multinational intensive care unit (ICU) network of registries of critically ill patients established in seven Asian low- and middle-income countries (LMICs). Methods: The Critical Care Asia federated registry platform enables ICUs to collect clinical, outcome and process data for aggregate and unit-level analysis. The evaluation used the standardised criteria of the Directory of Clinical Databases (DoCDat) and a framework for data quality assurance in medical registries. Six reviewers assessed structure, coverage, reliability and validity of the ICU registry data. Case mix and process measures on patient episodes from June to December 2020 were analysed. Results: Data on 20,507 consecutive patient episodes from 97 ICUs in Afghanistan, Bangladesh, India, Malaysia, Nepal, Pakistan and Vietnam were included. The quality level achieved according to the ten prespecified DoCDat criteria was high (average score 3.4 out of 4) as was the structural and organizational performance -- comparable to ICU registries in high-income countries. Identified strengths were types of variables included, reliability of coding, data completeness and validation. Potential improvements included extension of national coverage, optimization of recruitment completeness validation in all centers and the use of interobserver reliability checks. Conclusions: The Critical Care Asia platform evaluates well using standardised frameworks for data quality and equally to registries in resource-rich settings.

Integrating Electronic Medical Records and Claims Data for Influenza Vaccine Research

Real-world evidence (RWE) increasingly informs public health and healthcare decisions worldwide. A large database has been created (“Integrated Dataset”) that integrates primary care electronic medical records with pharmacy and medical claims data on >123 million US patients since 2014. This article describes the components of the Integrated Dataset and evaluates its representativeness to the US population and its potential use in evaluating influenza vaccine effectiveness. Representativeness to the US population (2014−2019) was evaluated by comparison with demographic information from the 2019 US census and the National Ambulatory Medical Care Survey (NAMCS). Variables included in the Integrated Dataset were evaluated against World Health Organization (WHO) defined key and non-critical variables for evaluating influenza vaccine performance. The Integrated Dataset contains a variety of information, including demographic data, patient medical history, diagnoses, immunizations, and prescriptions. Distributions of most age categories and sex were comparable with the US Census and NAMCS populations. The Integrated Dataset was less diverse by race and ethnicity. Additionally, WHO key and non-critical variables for the estimation of influenza vaccine effectiveness are available in the Integrated Dataset. In summary, the Integrated Dataset is generally representative of the US population and contains key variables for the assessment of influenza vaccine effectiveness.

Postlaunch evidence-generation studies for medical devices in Spain: the RedETS approach to integrate real-world evidence into decision making

The Monitoring Studies (MS) program, the approach developed by RedETS to generate postlaunch real-world evidence (RWE), is intended to complement and enhance the conventional health technology assessment process to support health policy decision making in Spain, besides informing other interested stakeholders, including clinicians and patients. The MS program is focused on specific uncertainties about the real effect, safety, costs, and routine use of new and insufficiently assessed relevant medical devices carefully selected to ensure the value of the additional research needed, by means of structured, controlled, participative, and transparent procedures. However, despite a clear political commitment and economic support from national and regional health authorities, several difficulties were identified along the development and implementation of the first wave of MS, delaying its execution and final reporting. Resolution of these difficulties at the regional and national levels and a greater collaborative impulse in the European Union, given the availability of an appropriate methodological framework already provided by EUnetHTA, might provide a faster and more efficient comparative RWE of improved quality and reliability at the national and international levels.

Validation process of a high-resolution database in a paediatric intensive care unit-Describing the perpetual patient's validation

RATIONALE

High data quality is essential to ensure the validity of clinical and research inferences based on it. However, these data quality assessments are often missing even though these data are used in daily practice and research.

AIMS AND OBJECTIVES

Our objective was to evaluate the data quality of our high-resolution electronic database (HRDB) implemented in our paediatric intensive care unit (PICU).

METHODS

We conducted a prospective validation study of a HRDB in a 32-bed paediatric medical, surgical, and cardiac PICU in a tertiary care freestanding maternal-child health centre in Canada. All patients admitted to the PICU with at least one vital sign monitored using a cardiorespiratory monitor connected to the central monitoring station.

RESULTS

Between June 2017 and August 2018, data from 295 patient days were recorded from medical devices and 4645 data points were video recorded and compared to the corresponding data collected in the HRDB. Statistical analysis showed an excellent overall correlation (R² = 1), accuracy (100%), agreement (bias = 0, limits of agreement = 0), completeness (2% missing data), and reliability (ICC = 1) between recorded and collected data within clinically significant pre-defined limits of agreement. Divergent points could all be explained.

CONCLUSIONS

This prospective validation of a representative sample showed an excellent overall data quality.

Hashmi M, Beane A, Murthy S, Dondorp AM, Haniffa R. Leveraging a Cloud-Based Critical Care Registry for COVID-19 Pandemic Surveillance and Research in Low- and Middle-Income Countries

The COVID-19 pandemic has revealed limitations in real-time surveillance needed for responsive health care action in low- and middle-income countries (LMICs). The Pakistan Registry for Intensive CarE (PRICE) was adapted to enable International Severe Acute Respiratory and emerging Infections Consortium (ISARIC)–compliant real-time reporting of severe acute respiratory infection (SARI). The cloud-based common data model and standardized nomenclature of the registry platform ensure interoperability of data and reporting between regional and global stakeholders. Inbuilt analytics enable stakeholders to visualize individual and aggregate epidemiological, clinical, and operational data in real time. The PRICE system operates in 5 of 7 administrative regions of Pakistan. The same platform supports acute and critical care registries in eleven countries in South Asia and sub-Saharan Africa. ISARIC-compliant SARI reporting was successfully implemented by leveraging the existing PRICE infrastructure in all 49 member intensive care units (ICUs), enabling clinicians, operational leads, and established stakeholders with responsibilities for coordinating the pandemic response to access real-time information on suspected and confirmed COVID-19 cases (N=592 as of May 2020) via secure registry portals. ICU occupancy rates, use of ICU resources, mechanical ventilation, renal replacement therapy, and ICU outcomes were reported through registry dashboards. This information has facilitated coordination of critical care resources, health care worker training, and discussions on treatment strategies. The PRICE network is now being recruited to international multicenter clinical trials regarding COVID-19 management, leveraging the registry platform. Systematic and standardized reporting of SARI is feasible in LMICs. Existing registry platforms can be adapted for pandemic research, surveillance, and resource planning.

Developing Registries and Effective Care Models for the Management of Hip Fractures: Aligning Political, Organizational Drivers with Clinical Outcomes

PURPOSE OF REVIEW

The aim of the review is to conduct a literature search on cost-effectiveness or cost savings of osteoporosis fracture liaison services.

RECENT FINDINGS

We identified four types of FLS. A total of 11 cost-effectiveness studies examining 15 models of secondary fracture prevention models were identified. Nine models were found to be cost-saving, and five were found to be cost-effective. It is possible to adopt a cost-effective model for fracture liaison services and expand across geographical regions. Adopting registries can have the added benefit of monitoring quality improvement practices and treatment outcomes. Challenges exist in implementing registries where centralized data collections across different chronic conditions are politically driving agendas. In order to align political and organizational strategic plans, a core set of outcome evaluations that are both focused on patient and provider experience in addition to treatment outcomes can be a step toward achieving better health and services.

A novel method to assess data quality in large medical registries and databases

BACKGROUND

There is no gold standard to assess data quality in large medical registries. Data auditing may be impeded by data protection regulations.

OBJECTIVE

To explore the applicability and usefulness of funnel plots as a novel tool for data quality control in critical care registries.

METHOD

The Swiss ICU-Registry from all 77 certified adult Swiss ICUs (2014 and 2015) was subjected to quality assessment (completeness/accuracy). For the analysis of accuracy, a list of logical rules and cross-checks was developed. Type and number of errors (true coding errors or implausible data) were calculated for each ICU, along with noticeable error rates (>mean + 3 SD in the variable's summary measure, or >99.8% CI in the respective funnel-plot).

RESULTS

We investigated 164 415 patient records with 31 items each (37 items: trauma diagnosis). Data completeness was excellent; trauma was the only incomplete item in 1495 of 9871 records (0.1%, 0.0%-0.6% [median, IQR]). In 15 572 patients records (9.5%), we found 3121 coding errors and 31 265 implausible situations; the latter primarily due to non-specific information on patients' provenance/diagnosis or supposed incoherence between diagnosis and treatments. Together, the error rate was 7.6% (5.9%-11%; median, IQR).

CONCLUSIONS

The Swiss ICU-Registry is almost complete and data quality seems to be adequate. We propose funnel plots as suitable, easy to implement instrument to assist in quality assurance of such a registry. Based on our analysis, specific feedback to ICUs with special-cause variation is possible and may promote such ICUs to improve the quality of their data.

Qualitative subjective assessment of a high-resolution database in a paediatric intensive care unit-Elaborating the perpetual patient's ID card

OBJECTIVE

The main purpose of our study was to subjectively assess the quality of a paediatric intensive care unit (PICU) database according to the Directory of Clinical Databases (DoCDat) criteria.

DESIGN AND SETTING

A survey was conducted between April 1 and June 15, 2018, among the Sainte Justine PICU research group.

POPULATION

Every member of this group whose research activity required the use of the database and/or who was involved in the development/validation of the database.

INTERVENTIONS

None.

MEASUREMENTS AND MAIN RESULTS

All 10 research team members (one Information Technology specialist, one junior medical student, and eight clinician researchers) who used the high-resolution database fulfilled the survey (100% response rate). The median quality level of the Sainte Justine PICU database across all the 10 criteria was 3 (2-4), rated on a 1 (worst) to 4 (best) numeric scale. When compared with previously assessed databases through the DoCDat criteria, we found that the Sainte Justine PICU database performance was similar.

CONCLUSIONS

The PICU high-resolution database appeared of good quality when subjectively assessed by the DoCDat criteria. Further validation procedures are mandatory. We suggest that data quality assessment and validation procedures should be reported when creating a new database.

The Japanese Intensive care PAtient Database (JIPAD): A national intensive care unit registry in Japan

PURPOSE

The Japanese Intensive care PAtient Database (JIPAD) was established to construct a high-quality Japanese intensive care unit (ICU) database.

MATERIALS AND METHODS

A data collection structure for consecutive ICU admissions in adults (≥16 years) and children (≤15 years) has been established in Japan since 2014. We herein report a current summary of the data in JIPAD for admissions between April 2015 and March 2017.

RESULTS

There were 21,617 ICU admissions from 21 ICUs (217 beds) including 8416 (38.9%) for postoperative or procedural monitoring, defined as adult admissions following elective surgery or for procedures and discharged alive within 24 h, 11,755 (54.4%) critically ill adults other than monitoring, and 1446 (6.7%) children. The standardized mortality ratios (SMRs) based on the Acute Physiology and Chronic Health Evaluation (APACHE) III-j, APACHE II, and Simplified Acute Physiology Score II scores in adults ranged from 0.387 to 0.534, whereas the SMR based on the Paediatric Index of Mortality 2 in children was 0.867.

CONCLUSION

The data revealed that the SMRs based on general severity scores in adults were low because of high proportions of elective and monitoring admission. The development of a new mortality prediction model for Japanese ICU patients is needed.

Real-world data: towards achieving the achievable in cancer care

The use of data from the real world to address clinical and policy-relevant questions that cannot be answered using data from clinical trials is garnering increased interest. Indeed, data from cancer registries and linked treatment records can provide unique insights into patients, treatments and outcomes in routine oncology practice. In this Review, we explore the quality of real-world data (RWD), provide a framework for the use of RWD and draw attention to the methodological pitfalls inherent to using RWD in studies of comparative effectiveness. Randomized controlled trials and RWD remain complementary forms of medical evidence; studies using RWD should not be used as substitutes for clinical trials. The comparison of outcomes between nonrandomized groups of patients who have received different treatments in routine practice remains problematic. Accordingly, comparative effectiveness studies need to be designed and interpreted very carefully. With due diligence, RWD can be used to identify and close gaps in health care, offering the potential for short-term improvement in health-care systems by enabling them to achieve the achievable.

Risk of sepsis in patients with primary aldosteronism

BACKGROUND

The interaction between hyperaldosteronism and immune dysfunction has been reported and glucocorticoid co-secretion is frequently found in primary aldosteronism (PA). The aforementioned conditions raise the possibility of the infection risk; however, clinical episodes of sepsis have not been reported in PA.

METHODS

Using Taiwan's National Health Insurance Research Database between 1997 and 2009, we identified PA and aldosterone-producing adenoma (APA) matched with essential hypertension (EH) at a 1:1 ratio by propensity scores. The incidences of sepsis and mortality after the index date were evaluated, and the risk factors of outcomes were identified using adjusted Cox proportional hazards models and taking mortality as a competing risk.

RESULTS

We enrolled 2448 patients with PA (male, 46.08%; mean age, 48.4 years). There were 875 patients who could be ascertained as APA. Taking mortality as the competing risk, APA patients had a lower incidence of sepsis than their matched EH patients (hazard ratio (HR) 0.29; P < 0.001) after target treatments. Patients receiving adrenalectomy showed a benefit of decreasing the risk of sepsis (PA vs EH, HR 0.14, P = 0.001; APA vs EH, HR 0.16, P = 0.003), but mineralocorticoid receptor antagonist treatment may differ. Compared with matched control cohorts, patients with APA had a lower risk of all-cause mortality (PA, adjusted HR 0.84, P = 0.050; APA, adjusted HR 0.31, P < 0.001) after target treatments.

CONCLUSIONS

Our study demonstrated that patients with PA/APA who underwent adrenalectomy could attenuate the risk of sepsis compared with their matched EH patients. We further found that APA patients with target treatments could decrease all-cause mortality compared with EH patients.

Creating a High-Frequency Electronic Database in the PICU: The Perpetual Patient

OBJECTIVE

Our objective was to construct a prospective high-quality and high-frequency database combining patient therapeutics and clinical variables in real time, automatically fed by the information system and network architecture available through fully electronic charting in our PICU. The purpose of this article is to describe the data acquisition process from bedside to the research electronic database.

DESIGN

Descriptive report and analysis of a prospective database.

SETTING

A 24-bed PICU, medical ICU, surgical ICU, and cardiac ICU in a tertiary care free-standing maternal child health center in Canada.

PATIENTS

All patients less than 18 years old were included at admission to the PICU.

INTERVENTIONS

None.

MEASUREMENTS AND MAIN RESULTS

Between May 21, 2015, and December 31, 2016, 1,386 consecutive PICU stays from 1,194 patients were recorded in the database. Data were prospectively collected from admission to discharge, every 5 seconds from monitors and every 30 seconds from mechanical ventilators and infusion pumps. These data were linked to the patient's electronic medical record. The database total volume was 241 GB. The patients' median age was 2.0 years (interquartile range, 0.0-9.0). Data were available for all mechanically ventilated patients (n = 511; recorded duration, 77,678 hr), and respiratory failure was the most frequent reason for admission (n = 360). The complete pharmacologic profile was synched to database for all PICU stays. Following this implementation, a validation phase is in process and several research projects are ongoing using this high-fidelity database.

CONCLUSIONS

Using the existing bedside information system and network architecture of our PICU, we implemented an ongoing high-fidelity prospectively collected electronic database, preventing the continuous loss of scientific information. This offers the opportunity to develop research on clinical decision support systems and computational models of cardiorespiratory physiology for example.

Dementia registries around the globe and their applications: A systematic review

Patient registries are valuable tools helping to address significant challenges in research, care, and policy. Registries, well embedded in many fields of medicine and public health, are relatively new in dementia. This systematic review presents the current situation in regards to dementia registries worldwide. We identified 31 dementia registries operating on an international, national, or local level between 1986 and 2016. More than half of the registries aimed to conduct or facilitate research, including preclinical research registries and registries recruiting research volunteers. Other dementia registries collected epidemiological or quality of care data. We present evidence of practical and economic outcomes of registries for research, clinical practice and policy, and recommendations for future development. Global harmonization of recruitment methods and minimum data would facilitate international comparisons. Registries provide a positive return on investment; their establishment and maintenance require ongoing support by government, policy makers, research funding bodies, clinicians, and individuals with dementia and their caregivers.

Validation of multisource electronic health record data: an application to blood transfusion data

Background

Although data from electronic health records (EHR) are often used for research purposes, systematic validation of these data prior to their use is not standard practice. Existing validation frameworks discuss validity concepts without translating these into practical implementation steps or addressing the potential influence of linking multiple sources. Therefore we developed a practical approach for validating routinely collected data from multiple sources and to apply it to a blood transfusion data warehouse to evaluate the usability in practice.

Methods

The approach consists of identifying existing validation frameworks for EHR data or linked data, selecting validity concepts from these frameworks and establishing quantifiable validity outcomes for each concept. The approach distinguishes external validation concepts (e.g. concordance with external reports, previous literature and expert feedback) and internal consistency concepts which use expected associations within the dataset itself (e.g. completeness, uniformity and plausibility). In an example case, the selected concepts were applied to a transfusion dataset and specified in more detail.

Results

Application of the approach to a transfusion dataset resulted in a structured overview of data validity aspects. This allowed improvement of these aspects through further processing of the data and in some cases adjustment of the data extraction. For example, the proportion of transfused products that could not be linked to the corresponding issued products initially was 2.2% but could be improved by adjusting data extraction criteria to 0.17%.

Conclusions

This stepwise approach for validating linked multisource data provides a basis for evaluating data quality and enhancing interpretation. When the process of data validation is adopted more broadly, this contributes to increased transparency and greater reliability of research based on routinely collected electronic health records.

Electronic supplementary material

The online version of this article (doi:10.1186/s12911-017-0504-7) contains supplementary material, which is available to authorized users.

Elderly patients hospitalized in the ICU in France: a population-based study using secondary data from the national hospital discharge database

RATIONALE, AIMS AND OBJECTIVES

In the global context of population ageing, understanding and monitoring intensive care use by the elderly is a strategic issue. National-level data are needed to overcome sampling biases that often limit epidemiologic studies of the critically ill elderly. The objective of this study was to describe intensive care use for hospitalized elderly patients using secondary data from the French national hospital discharge database.

METHOD

Structured assessment of the national database coverage and accuracy; cross-sectional analysis of hospitalizations including at least one admission in an intensive care unit (ICU) for patients aged ≥ 80 years from 1 January to 31 December 2009.

RESULTS

In 2009, people aged ≥ 80 years accounted for 5.4% of the population but 15.3% of the 215 210 adult hospitalizations involving intensive care in France. In this elderly group, the mean age was 84.0 (± 3.56) years, and 51.6% were male. In-hospital mortality was 33.9%. The median time spent in the ICU was 3 [interquartile range (IQR), 2-8] days, the median time spent in hospital was 14 (IQR, 8-24) days and 9% of hospitalizations ended by the patient's death involved intensive care. A surgical procedure was included in 43% of hospitalizations. Medical and surgical diagnosis-related group hospitalizations were characterized by significant differences in volume, mortality, ICU days and costs.

CONCLUSIONS

There was marked clinical heterogeneity in the population of elderly patients hospitalized in the ICU. These data provide baseline information and prompt further studies comparing intensive care utilization across age groups, between countries and over time.

Machine Learning and Decision Support in Critical Care

Clinical data management systems typically provide caregiver teams with useful information, derived from large, sometimes highly heterogeneous, data sources that are often changing dynamically. Over the last decade there has been a significant surge in interest in using these data sources, from simply re-using the standard clinical databases for event prediction or decision support, to including dynamic and patient-specific information into clinical monitoring and prediction problems. However, in most cases, commercial clinical databases have been designed to document clinical activity for reporting, liability and billing reasons, rather than for developing new algorithms. With increasing excitement surrounding "secondary use of medical records" and "Big Data" analytics, it is important to understand the limitations of current databases and what needs to change in order to enter an era of "precision medicine." This review article covers many of the issues involved in the collection and preprocessing of critical care data. The three challenges in critical care are considered: compartmentalization, corruption, and complexity. A range of applications addressing these issues are covered, including the modernization of static acuity scoring; on-line patient tracking; personalized prediction and risk assessment; artifact detection; state estimation; and incorporation of multimodal data sources such as genomic and free text data.

Making quality registers supporting improvements: a systematic review of the data visualization in 5 quality registries

Traditionally, quality registries have been initiated, developed, and used by physicians essentially for research purposes. There is an unrealized opportunity to expand and strengthen the contribution of quality registries in health care quality improvement. This article aims to characterize quality registry annual reports regarding factors deemed important to process improvement. The 2012 annual reports of the 5 most highly developed Swedish quality registries were examined. Each of the 636 charts included was coded according to an abstraction form. Results show that league tables are highly prevalent, whereas funnel plots and control charts are rare. Health care quality is monitored over time on the basis of few and highly aggregated measurements, and it is usually measured using percentages. In conclusion, quality registry annual reports lack both the level of detail and the consideration of random variation necessary to being able to be systematically used in process improvement. Users of annual reports are recommended caution when discussing differences in quality, both over time and across health care providers, as they can be due to chance and insufficient guidance is provided on the reports in this regard. To better support process improvement, annual reports should thus be more detailed and give more consideration to random variation.

The VASOGRADE: A Simple Grading Scale for Prediction of Delayed Cerebral Ischemia After Subarachnoid Hemorrhage

BACKGROUND AND PURPOSE

Patients are classically at risk of delayed cerebral ischemia (DCI) after aneurysmal subarachnoid hemorrhage. We validated a grading scale-the VASOGRADE-for prediction of DCI.

METHODS

We used data of 3 phase II randomized clinical trials and a single hospital series to assess the relationship between the VASOGRADE and DCI. The VASOGRADE derived from previously published risk charts and consists of 3 categories: VASOGRADE-Green (modified Fisher scale 1 or 2 and World Federation of Neurosurgical Societies scale [WFNS] 1 or 2); VASOGRADE-Yellow (modified Fisher 3 or 4 and WFNS 1-3); and VASOGRADE-Red (WFNS 4 or 5, irrespective of modified Fisher grade). The relation between the VASOGRADE and DCI was assessed by logistic regression models. The predictive accuracy of the VASOGRADE was assessed by receiver operating characteristics curve and calibration plots.

RESULTS

In a cohort of 746 patients, the VASOGRADE significantly predicted DCI (P<0.001). The VASOGRADE-Yellow had a tendency for increased risk for DCI (odds ratio [OR], 1.31; 95% CI, 0.77-2.23) when compared with VASOGRADE-Green; those with VASOGRADE-Red had a 3-fold higher risk of DCI (OR, 3.19; 95% CI, 2.07-4.50). Studies were not a significant confounding factor between the VASOGRADE and DCI. The VASOGRADE had an adequate discrimination for prediction of DCI (area under the receiver operating characteristics curve=0.63) and good calibration.

CONCLUSIONS

The VASOGRADE results validated previously published risk charts in a large and diverse sample of subarachnoid hemorrhage patients, which allows DCI risk stratification on presentation after subarachnoid hemorrhage. It could help to select patients at high risk of DCI, as well as standardize treatment protocols and research studies.

The rise of big clinical databases

BACKGROUND

The routine collection of large amounts of clinical data, 'big data', is becoming more common, as are research studies that make use of these data source. The aim of this paper is to provide an overview of the uses of data from large multi-institution clinical databases for research.

METHODS

This article considers the potential benefits, the types of data source, and the use to which the data is put. Additionally, the main challenges associated with using these data sources for research purposes are considered.

RESULTS

Common uses of the data include: providing population characteristics; identifying risk factors and developing prediction (diagnostic or prognostic) models; observational studies comparing different interventions; exploring variation between healthcare providers; and as a supplementary source of data for another study. The main advantages of using such big data sources are their comprehensive nature, the relatively large number of patients they comprise, and the ability to compare healthcare providers. The main challenges are demonstrating data quality and confidently applying a causal interpretation to the study findings.

CONCLUSION

Large clinical database research studies are becoming ubiquitous and offer a number of potential benefits. However, the limitations of such data sources must not be overlooked; each research study needs to be considered carefully in its own right, together with the justification for using the data for that specific purpose.

Establishment of a regional Danish database for patients with a stoma

AIM

To present the Danish Stoma Database Capital Region with clinical variables related to stoma creation including colostomy, ileostomy and urostomy.

METHOD

The stomatherapists in the Capital Region of Denmark developed a database covering patient identifiers, interventions, conditions, short-term outcome, long-term outcome and known major confounders. The completeness of data was validated against the Danish National Patient Register.

RESULTS

In 2013, five hospitals included data from 1123 patients who were registered during the year. The types of stomas formed from 2007 to 2013 showed a variation reflecting the subspecialization and surgical techniques in the centres. Between 92 and 94% of patients agreed to participate in the standard programme aimed at handling of the stoma and more than 88% of patients having planned surgery had the stoma site marked pre-operatively.

CONCLUSION

The database is fully operational with high data completeness and with data about patients with a stoma from before surgery up to 12 months after surgery. The database provides a solid basis for professional learning, clinical research and benchmarking.

Use of Electronic Health-Related Datasets in Nursing and Health-Related Research

Datasets of gigabyte size are common in medical sciences. There is increasing consensus that significant untapped knowledge lies hidden in these large datasets. This review article aims to discuss Electronic Health-Related Datasets (EHRDs) in terms of types, features, advantages, limitations, and possible use in nursing and health-related research. Major scientific databases, MEDLINE, ScienceDirect, and Scopus, were searched for studies or review articles regarding using EHRDs in research. A total number of 442 articles were located. After application of study inclusion criteria, 113 articles were included in the final review. EHRDs were categorized into Electronic Administrative Health-Related Datasets and Electronic Clinical Health-Related Datasets. Subcategories of each major category were identified. EHRDs are invaluable assets for nursing the health-related research. Advanced research skills such as using analytical softwares, advanced statistical procedures, dealing with missing data and missing variables will maximize the efficient utilization of EHRDs in research.

Improved prognosis of septic shock in patients with cirrhosis: a multicenter study*

OBJECTIVE

To determine the evolution of the outcome of patients with cirrhosis and septic shock.

DESIGN

A 13-year (1998-2010) multicenter retrospective cohort study of prospectively collected data.

SETTING

The Collège des Utilisateurs des Bases des données en Réanimation (CUB-Réa) database recording data related to admissions in 32 ICUs in Paris area.

PATIENTS

Thirty-one thousand two hundred fifty-one patients with septic shock were analyzed; 2,383 (7.6%) had cirrhosis.

INTERVENTIONS

None.

MEASUREMENTS AND MAIN RESULTS

Compared with noncirrhotic patients, patients with cirrhosis had higher Simplified Acute Physiology Score II (63.1 ± 22.7 vs 58.5 ± 22.8, p < 0.0001) and higher prevalence of renal (71.5% vs 54.8%, p < 0.0001) and neurological (26.1% vs 19.5%, p < 0.0001) dysfunctions. Over the study period, in-ICU and in-hospital mortality was higher in patients with cirrhosis (70.1% and 74.5%) compared with noncirrhotic patients (48.3% and 51.7%, p < 0.0001 for both comparisons). Cirrhosis was independently associated with an increased risk of death in ICU (adjusted odds ratio = 2.524 [2.279-2.795]). In patients with cirrhosis, factors independently associated with in-ICU mortality were as follows: admission for a medical reason, Simplified Acute Physiology Score II, mechanical ventilation, renal replacement therapy, spontaneous bacterial peritonitis, positive blood culture, and infection by fungus, whereas direct admission and admission during the most recent midterm period (2004-2010) were associated with a decreased risk of death. From 1998 to 2010, prevalence of septic shock in patients with cirrhosis increased from 8.64 to 15.67 per 1,000 admissions to ICU (p < 0.0001) and their in-ICU mortality decreased from 73.8% to 65.5% (p = 0.01) despite increasing Simplified Acute Physiology Score II. In-ICU mortality decreased from 84.7% to 68.5% for those patients placed under mechanical ventilation (p = 0.004) and from 91.2% to 78.4% for those who received renal replacement therapy (p = 0.04).

CONCLUSIONS

The outcome of patients with cirrhosis and septic shock has markedly improved over time, akin to the noncirrhotic population. In 2010, the in-ICU survival rate was 35%, which now fully justifies to admit these patients to ICU.

Feasibility of using administrative data to compare hospital performance in the EU

OBJECTIVE

To describe hospitals' organizational arrangements relevant to the abstraction of administrative data, to report on the completeness of administrative data collected and to assess associations between organizational arrangements and completeness of data submission.

DESIGN

A cross-sectional

STUDY DESIGN

utilizing administrative data.

SETTING AND PARTICIPANTS

Randomly selected hospitals from seven European countries (The Czech Republic, France, Germany, Poland, Portugal, Spain, and Turkey).

MAIN OUTCOME MEASURES

Completeness of data submission for four quality indicators: mortality after acute myocardial infarction, stroke and hip fractures and complications after normal delivery.

RESULTS

In general, hospitals were able to produce data on the four indicators required for this research study. A substantial proportion had missing data on one or more data items. The proportion of hospitals that was able to produce more detailed indicators of relevance for quality monitoring and improvement was low and ranged from 40.1% for thrombolysis performed on patients with acute ischemic stroke to 63.8% for hip-fracture operations performed within 48 h after admission for patients aged 65 or older. National factors were strong predictors of data completeness on the studied indicators.

CONCLUSIONS

At present, hospital administrative databases do not seem to be an appropriate source of information for comparison of hospital performance across the countries of the EU. However, given that this is a dynamic field, changes to administrative databases may make this possible in the near future. Such changes could be accelerated by an in-depth comparative analysis of the issues of using administrative data for comparisons of hospital performances in EU countries.

Existing data analysis in pediatric critical care research

Our objectives were to review and categorize the existing data sources that are important to pediatric critical care medicine (PCCM) investigators and the types of questions that have been or could be studied with each data source. We conducted a narrative review of the medical literature, categorized the data sources available to PCCM investigators, and created an online data source registry. We found that many data sources are available for research in PCCM. To date, PCCM investigators have most often relied on pediatric critical care registries and treatment- or disease-specific registries. The available data sources vary widely in the level of clinical detail and the types of questions they can reliably answer. Linkage of data sources can expand the types of questions that a data source can be used to study. Careful matching of the scientific question to the best available data source or linked data sources is necessary. In addition, rigorous application of the best available analysis techniques and reporting consistent with observational research standards will maximize the quality of research using existing data in PCCM.

Using existing data to address important clinical questions in critical care

OBJECTIVE

With important technological advances in healthcare delivery and the Internet, clinicians and scientists now have access to overwhelming number of available databases capturing patients with critical illness. Yet, investigators seeking to answer important clinical or research questions with existing data have few resources that adequately describe the available sources and the strengths and limitations of each. This article reviews an approach to selecting a database to address health services and outcomes research questions in critical care, examines several databases that are commonly used for this purpose, and briefly describes some strengths and limitations of each.

DATA SOURCES

Narrative review of the medical literature.

SUMMARY

The available databases that collect information on critically ill patients are numerous and vary in the types of questions they can optimally answer. Selection of a data source must consider not only accessibility but also the quality of the data contained within the database, and the extent to which it captures the necessary variables for the research question. Questions seeking causal associations (e.g., effect of treatment on mortality) usually either require secondary data that contain detailed information about demographics, laboratories, and physiology to best address nonrandom selection or sophisticated study design. Purely descriptive questions (e.g., incidence of respiratory failure) can often be addressed using secondary data with less detail such as administrative claims. Although each database has its own inherent limitations, all secondary analyses will be subject to the same challenges of appropriate study design and good observational research.

CONCLUSION

The literature demonstrates that secondary analyses can have significant impact on critical care practice. While selection of the optimal database for a particular question is a necessary part of high-quality analyses, it is not sufficient to guarantee an unbiased study. Thoughtful and well-constructed study design and analysis approaches remain equally important pillars of robust science. Only through responsible use of existing data will investigators ensure that their study has the greatest impact on critical care practice and outcomes.

Use of national clinical databases for informing and for evaluating health care policies

Policy-makers and analysts could make use of national clinical databases either to inform or to evaluate meso-level (organisation and delivery of health care) and macro-level (national) policies. Reviewing the use of 15 of the best established databases in England, we identify and describe four published examples of each use. These show that policy-makers can either make use of the data itself or of research based on the database. For evaluating policies, the major advantages are the huge sample sizes available, the generalisability of the data, its immediate availability and historic information. The principal methodological challenges involve the need for risk adjustment and time-series analysis. Given their usefulness in the policy arena, there are several reasons why national clinical databases have not been used more, some due to a lack of 'push' by their custodians and some to the lack of 'pull' by policy-makers. Greater exploitation of these valuable resources would be facilitated by policy-makers' and custodians' increased awareness, minimisation of legal restrictions on data use, improvements in the quality of databases and a library of examples of applications to policy.

The Nottingham Head Injury Register: A Survey of 1,276 Adult Cases of Moderate and Severe Traumatic Brain Injury in a British Neurosurgery Centre

The Nottingham Head Injury Register was established as an audit and research tool for traumatic brain injury (TBI). Study aims included examination of the epidemiology of UK TBI and identification of factors improving management and outcome. Moderate and severe TBI patients admitted to a UK neurosurgery centre were recruited over a 10-year period from 1993 to 2003. A dataset covering all aspects of TBI, including demography, clinical features, investigations, management, complications and outcome were captured prospectively into a computerised database. The study group comprised 1,276 adult patients aged more than 16 years old. The median age of subjects was 35 years, with 78% of the population consisting of male patients. Road traffic accidents were responsible for 49% of injuries; 37% of the study population had moderate TBI and 63% severe TBI. CT brain scan was performed in 92% of cases showing a mass lesion in 41% and diffuse lesion in 59%. Traumatic subarachnoid haemorrhage was present in 34%. At 12 months 32% of patients had died, 13% were severely disabled, 22% were moderately disabled and 33% had made a good recovery. Findings were comparable with other previous large series of TBI patients from international studies. Registry data is useful to define TBI epidemiology and management, and in validation of prognostic scoring systems.

Data linkage: a powerful research tool with potential problems

BACKGROUND

Policy makers, clinicians and researchers are demonstrating increasing interest in using data linked from multiple sources to support measurement of clinical performance and patient health outcomes. However, the utility of data linkage may be compromised by sub-optimal or incomplete linkage, leading to systematic bias. In this study, we synthesize the evidence identifying participant or population characteristics that can influence the validity and completeness of data linkage and may be associated with systematic bias in reported outcomes.

METHODS

A narrative review, using structured search methods was undertaken. Key words "data linkage" and Mesh term "medical record linkage" were applied to Medline, EMBASE and CINAHL databases between 1991 and 2007. Abstract inclusion criteria were; the article attempted an empirical evaluation of methodological issues relating to data linkage and reported on patient characteristics, the study design included analysis of matched versus unmatched records, and the report was in English. Included articles were grouped thematically according to patient characteristics that were compared between matched and unmatched records.

RESULTS

The search identified 1810 articles of which 33 (1.8%) met inclusion criteria. There was marked heterogeneity in study methods and factors investigated. Characteristics that were unevenly distributed among matched and unmatched records were; age (72% of studies), sex (50% of studies), race (64% of studies), geographical/hospital site (93% of studies), socio-economic status (82% of studies) and health status (72% of studies).

CONCLUSION

A number of relevant patient or population factors may be associated with incomplete data linkage resulting in systematic bias in reported clinical outcomes. Readers should consider these factors in interpreting the reported results of data linkage studies.

The use of databases and registries to enhance colonoscopy quality

Administrative databases, registries, and clinical databases are designed for different purposes and therefore have different advantages and disadvantages in providing data for enhancing quality. Administrative databases provide the advantages of size, availability, and generalizability, but are subject to constraints inherent in the coding systems used and from data collection methods optimized for billing. Registries are designed for research and quality reporting but require significant investment from participants for secondary data collection and quality control. Electronic health records contain all of the data needed for quality research and measurement, but that data is too often locked in narrative text and unavailable for analysis. National mandates for electronic health record implementation and functionality will likely change this landscape in the near future.

Comparative indicators for cancer network management in England: availability, characteristics and presentation

Background

In 2000, the national cancer plan for England created 34 cancer networks, new organisational structures to coordinate services across populations varying between a half and three million people. We investigated the availability of data sets reflecting measures of structure, process and outcome that could be used to support network management.

Methods

We investigated the properties of national data sets relating to four common cancers – breast, colorectal, lung and prostate. We reviewed the availability and completeness of these data sets, identified leading items within each set and put them into tables of the 34 cancer networks. We also investigated methods of presentation.

Results

The Acute Hospitals Portfolio and the Cancer Standards Peer Review recorded structural characteristics at hospital and cancer service level. Process measures included Hospital Episode Statistics, recording admissions, and Hospital Waiting-List data. Patient outcome measures included the National Survey of Patient Satisfaction for cancer, and cancer survival, drawn from cancer registration. Data were drawn together to provide an exemplar indicator set a single network, and methods of graphical presentation were considered.

Conclusion

While not as yet used together in practice, comparative indicators are available within the National Health Service in England for use in performance assessment by cancer networks.

Databases as policy instruments. About extending networks as evidence-based policy

Background

This article seeks to identify the role of databases in health policy. Access to information and communication technologies has changed traditional relationships between the state and professionals, creating new systems of surveillance and control. As a result, databases may have a profound effect on controlling clinical practice.

Methods

We conducted three case studies to reconstruct the development and use of databases as policy instruments. Each database was intended to be employed to control the use of one particular pharmaceutical in the Netherlands (growth hormone, antiretroviral drugs for HIV and Taxol, respectively). We studied the archives of the Dutch Health Insurance Board, conducted in-depth interviews with key informants and organized two focus groups, all focused on the use of databases both in policy circles and in clinical practice.

Results

Our results demonstrate that policy makers hardly used the databases, neither for cost control nor for quality assurance. Further analysis revealed that these databases facilitated self-regulation and quality assurance by (national) bodies of professionals, resulting in restrictive prescription behavior amongst physicians.

Conclusion

The databases fulfill control functions that were formerly located within the policy realm. The databases facilitate collaboration between policy makers and physicians, since they enable quality assurance by professionals. Delegating regulatory authority downwards into a network of physicians who control the use of pharmaceuticals seems to be a good alternative for centralized control on the basis of monitoring data.

Clinical databases in physical therapy

Clinical databases in physical therapy provide increasing opportunities for research into physical therapy theory and practice. At present, information on the characteristics of existing databases is lacking. The purpose of this study was to identify clinical databases in which physical therapists record data on their patients and treatments and to investigate the basic aspects, data sets, output, management, and data quality of the databases. Identification of the databases was performed by contacting members of the World Confederation for Physical Therapy, searching Pubmed, searching the Internet, and snowball sampling. A structured questionnaire was used to study the characteristics of the databases. The search was restricted to North America, Australia, Israel, and Western Europe. Seven clinical databases on physical therapy were identified. Four databases collected data on specific patient categories, whereas the others collected data on all patients. All databases collected data on patient characteristics, referrals, diagnoses, treatments, and closure, whereas some databases also collected functional status information. The purposes of the databases were diverse, but they can be summarized as quality improvement, research, and performance management. Although clinical databases are new to the field, they offer great potential for physical therapy research. Potential can be increased by further cooperation among databases allowing international comparative studies.

Why is there such a difference in outcome between Australian intensive care units and others?

PURPOSE OF REVIEW

The aim of this article is to assess the data on clinical outcomes for critically ill patients admitted to Australian and New Zealand intensive care units in comparison to information available for similar patients in other counties

RECENT FINDINGS

Australia and New Zealand have been collecting standardized data intensive care unit admissions for over a decade. The Australian and New Zealand Intensive Care Society Database Management Committee has developed a high quality database of close to 600 000 adult intensive care unit admissions. Although comparisons suffer from significant methodological, case-mix and process differences, which make their findings easily subject to criticism, interrogation of this database and of data from clusters of intensive care units within this system consistently yields patient outcomes, which are better than outcomes reported from other nations or international studies for similar patients. In addition, Australia and New Zealand has now achieved the highest rate of patient enrollment in an investigator-initiated multicentre randomized controlled trials.

SUMMARY

Although comparisons in outcome between Australia and New Zealand intensive care units and other units worldwide may not have sufficient scientific rigour to truly reflect better national outcomes, many features of Australian and New Zealand units are unique and worthy of consideration by other national systems as they consider their strategic national goals for the next decade.

Systematic review and evaluation of physiological track and trigger warning systems for identifying at-risk patients on the ward

OBJECTIVE

Physiological track and trigger warning systems (TTs) are used to identify patients outside critical care areas at risk of deterioration and to alert a senior clinician, Critical Care Outreach Service, or equivalent. The aims of this work were: to describe published TTs and the extent to which each has been developed according to established procedures; to review the published evidence and available data on the reliability, validity and utility of existing systems; and to identify the best TT for timely recognition of critically ill patients.

DESIGN AND SETTING

Systematic review of studies identified from electronic, citation and hand searching, and expert informants. Cohort study of data from 31 acute hospitals in England and Wales.

MEASUREMENTS AND RESULTS

Thirty-six papers were identified describing 25 distinct TTs. Thirty-one papers described the use of a TT, and five were studies examining the development or testing of TTs. None of the studies met all methodological quality standards. For the cohort study, outcome was measured by a composite of death, admission to critical care, 'do not attempt resuscitation' or cardiopulmonary resuscitation. Fifteen datasets met pre-defined quality criteria. Sensitivities and positive predictive values were low, with median (quartiles) of 43.3 (25.4-69.2) and 36.7 (29.3-43.8), respectively.

CONCLUSION

A wide variety of TTs were in use, with little evidence of reliability, validity and utility. Sensitivity was poor, which might be due in part to the nature of the physiology monitored or to the choice of trigger threshold. Available data were insufficient to identify the best TT.

Effect of hospital ownership status and payment structure on the adoption and use of drug-eluting stents for percutaneous coronary interventions

BACKGROUND

The impact of the use of drug-eluting stents in percutaneous coronary intervention (PCI) on cardiac care is still uncertain. We examined the influence of systemic factors, such as hospital ownership status, organizational characteristics and payment structure, on the use of drug-eluting stents in PCI and the effect on cardiac surgery volume.

METHODS

We conducted a cross-sectional analysis of drug-eluting stent use in 12 993 patients undergoing PCI with stenting (drug-eluting or bare-metal) and time-series regression analyses of the monthly number of cardiac surgery and PCI procedures performed using data collected from 1998 to 2004 at 13 public and private hospitals in the Emilia-Romagna region of Italy.

RESULTS

Public hospitals used drug-eluting stents more selectively than private hospitals, targeting the new device to patients at high risk of adverse events. The time-series regression analyses showed that the number of PCI procedures performed per year increased during this period, both in public (slope coefficient 36.4, 95% confidence interval [CI] 30.2 to 43.1) and private centres (slope coefficient 6.4, 95% CI 3.1 to 9.2 ). Concurrently, there was a reduction in the number of isolated coronary artery bypass graft (CABG) surgeries, although the degree of change was higher in public than in private hospitals (coefficient -16.1 v. -6.2 respectively ). The number of CABG procedures associated with valve surgery decreased in public hospitals (coefficient -5.0, 95% CI -6.1 to -3.8) but increased in private hospitals (coefficient 4.1, 95% CI 2.0 to 6.1).

INTERPRETATION

Public and private hospitals behaved differently in adopting drug-eluting stents and in using PCI with drug-eluting stents as a substitute for surgical revascularization.

Development and implementation of a high-quality clinical database: the Australian and New Zealand Intensive Care Society Adult Patient Database

OBJECTIVE

To describe the development of a binational intensive care database.

SETTING

One hundred thirty-eight intensive care units (ICUs) in Australia and New Zealand.

METHODS

A structure was developed to enable ICUs to submit data for central and local analysis. Reports were developed to allow comparison with similar ICU types and against published mortality prediction models. The database was evaluated according to (a) the criteria of the Directory of Clinical Databases (DoCDat) and (b) a proposed framework for data quality assurance in medical registries.

RESULTS

Between January 1987 and December 2003, 444,147 data sets were collected from 121 (72.5%) of 167 Australian and 10 (37.0%) of 27 New Zealand ICUs. Data sets from more than 60000 ICU admissions were submitted in 2003. Overall hospital mortality was 14.5%. The mean quality level achieved according to DoCDat criteria was high as was performance against a proposed framework for data quality. The provision of no-cost software has been vitally important to the success of the database.

CONCLUSION

A high-quality ICU database has successfully been implemented in Australia and New Zealand and is now used as a routine quality assurance and peer review tool. Similar developments may be both possible and desirable in other countries.

Health services research in critical care using administrative data

The aim of health services research is to provide unbiased, scientific evidence to influence health services policy at all levels. Secondary analysis of administrative data can be employed for the purpose of evaluating questions relevant to health services research in the field of critical care. This article provides an overview of the topic and specifically reviews the key components to evaluating and performing research in critical care using administrative data, including how to evaluate the quality of administrative data itself, and also how to evaluate the quality of studies that employ administrative data. The article concludes with a step-wise approach to conducting studies of critical care based on administrative data.

Lessons from developing and running a clinical database for colorectal cancer

BACKGROUND

Recent policy developments in the UK require the routine monitoring of the performance of cancer services. Developing and using clinical databases is one approach to meet this objective, but to date their implementation has been challenging.

OBJECTIVE

To describe the development of the Thames Cancer Registry clinical database for colorectal cancer, and to present the lessons learnt in the first five years since its establishment.

METHODS

Planning of this clinical database began in 1998. Detailed variables for the data set were derived by analysis of national standards and guidelines. Structured pro formas were designed to abstract data from clinical notes. A pilot study over 12 months collected 400 cases from seven hospital trusts in one cancer network. Data collection over the wider North Thames area began in 1999.

RESULTS

The number of new records entered each year into the database rose from 747 in 1999 to 1107 in 2002. By 2004, it held a total of 8500. However, participation and completeness of data collection varied between trusts. Currently only 18 of 26 trusts in the area submit data and only 12 have done so every year. Overall completeness for key demographic and treatment variables has been between 80 and 100% but less so for more detailed diagnostic and treatment variables (40-60%). Barriers to implementation in trusts could be grouped as organizational, professional and data-related. Organizational barriers have included changes in the cancer networks, variability in trust commitment to different data sets and lack of personnel to enter data consistently. Professional barriers have included competing priorities and varying commitments within the multidisciplinary clinical teams. Data-related barriers include the wide range of database formats that are used in trusts, and a tendency for data to be collected at the end of the year rather than continuously.

CONCLUSIONS

Creating and maintaining a clinical database is a time-consuming and complex undertaking. Completeness of ascertainment and quality are major issues of concern. Key lessons from this project have been that the commitment of clinicians and the ability of trusts to provide consistent support for data collection are crucial.

SAPS�II revisited

OBJECTIVE

To construct and validate an update of the Simplified Acute Physiology Score II (SAPS II) for the evaluation of clinical performance of Intensive Care Units (ICU).

DESIGN AND SETTING

Retrospective analysis of prospectively collected multicenter data in 32 ICUs located in the Paris area belonging to the Cub-Rea database and participating in a performance evaluation project.

PATIENTS

33,471 patients treated between 1999 and 2000.

MEASUREMENTS AND RESULTS

Two logistic regression models based on SAPS II were developed to estimate in-hospital mortality among ICU patients. The second model comprised reevaluation of original items of SAPS II and integration of the preadmission location and chronic comorbidity. Internal and external validation were performed. In the two validation samples the most complex model had better calibration than the original SAPS II for in-hospital mortality but its discrimination was not significantly higher (area under ROC curve 0.89 vs. 0.87 for SAPS II). Second-level customization and integration of new items improved uniformity of fit for various categories of patients except for diagnosis-related groups. The rank order of ICUs was modified according to the model used.

CONCLUSIONS

The overall performance of SAPS II derived models was good, even in the context of a community cohort and routinely gathered data. However, one-half the variation of outcome remains unexplained after controlling for admission characteristics, and uniformity of prediction across diagnostic subgroups was not achieved. Differences in case-mix still limit comparisons of quality of care.