Collecting information on the quality of prescribing in primary care using semi-automatic data extraction from GPs’ electronic medical records

Extracting data from a chemotherapy prescription platform for real-world oncology research in the UK: a pilot study

Aim: In the UK, there are limited data sources for evaluating real-world research questions related to oncology therapy. We conducted a pilot study to investigate the feasibility of extracting data directly from a chemotherapy prescription platform (ChemoCare^®) utilized in standard care. Patients & methods: Concordance was compared with data extracted manually for patients with advanced melanoma as part of a concurrent chart review (gold-standard) using Cohen's kappa and the intraclass correlation coefficient. Results: There was high concordance between data automatically extracted from the prescription platform and chart review data. Conclusion: This pilot can be used as a framework for future studies using direct, automated extraction from prescription platforms.

Smart Extraction and Analysis System for Clinical Research

BACKGROUND

With the increasing use of electronic health records (EHRs), there is a growing need to expand the utilization of EHR data to support clinical research. The key challenge in achieving this goal is the unavailability of smart systems and methods to overcome the issue of data preparation, structuring, and sharing for smooth clinical research.

MATERIALS AND METHODS

We developed a robust analysis system called the smart extraction and analysis system (SEAS) that consists of two subsystems: (1) the information extraction system (IES), for extracting information from clinical documents, and (2) the survival analysis system (SAS), for a descriptive and predictive analysis to compile the survival statistics and predict the future chance of survivability. The IES subsystem is based on a novel permutation-based pattern recognition method that extracts information from unstructured clinical documents. Similarly, the SAS subsystem is based on a classification and regression tree (CART)-based prediction model for survival analysis.

RESULTS

SEAS is evaluated and validated on a real-world case study of head and neck cancer. The overall information extraction accuracy of the system for semistructured text is recorded at 99%, while that for unstructured text is 97%. Furthermore, the automated, unstructured information extraction has reduced the average time spent on manual data entry by 75%, without compromising the accuracy of the system. Moreover, around 88% of patients are found in a terminal or dead state for the highest clinical stage of disease (level IV). Similarly, there is an ∼36% probability of a patient being alive if at least one of the lifestyle risk factors was positive.

CONCLUSION

We presented our work on the development of SEAS to replace costly and time-consuming manual methods with smart automatic extraction of information and survival prediction methods. SEAS has reduced the time and energy of human resources spent unnecessarily on manual tasks.

Quality indicators as a tool in improving the introduction of new medicines

Quality indicators are increasingly used as a tool to achieve safe and quality clinical care, cost-effective therapy, for professional learning, remuneration, accreditation and financial incentives. A substantial number focus on drug therapy but few address the introduction of new medicines even though this is a burning issue. The objective was to describe the issues and challenges in designing and implementing a transparent indicator framework and evaluation protocol for the introduction of new medicines and to provide guidance on how to apply quality indicators in the managed entry of new medicines. Quality indicators need to be developed early to assess whether new medicines are introduced appropriately. A number of key factors need to be addressed when developing, applying and evaluating indicators including dimensions of quality, suggested testing protocols, potential data sources, key implementation factors such as intended and unintended consequences, budget impact and cost-effectiveness, assuring the involvement of the medical professions, patients and the public, and reliable and easy-to-use computerized tools for data collection and management. Transparent approaches include the need for any quality indicators developed to handle conflict of interests to enhance their validity and acceptance. The suggested framework and indicator testing protocol may be useful in assessing the applicability of indicators for new medicines and may be adapted to healthcare settings worldwide. The suggestions build on existing literature to create a field testing methodology that can be used to produce country-specific quality indicators for new medicines as well as a cross international approach to facilitate access to new medicines.

Reasons for encounters, investigations, referrals, diagnoses and treatments in general practice in Sweden--a multicentre pilot study using electronic patient records

OBJECTIVE

To investigate reasons for encounters, investigations, referrals, diagnoses and treatments in everyday general practice, using electronic patient records (EPR), and possible related differences concerning gender, socio-economic status (SES) and practice location.

METHOD

Four Swedish primary care centres using EPR participated. Distributions of symptoms, investigations, diagnoses and prescribed drugs were registered.

RESULTS

In 1055 encounters, the mean patient age was 53; 59% were women. The most common reasons for the encounter were musculoskeletal (21.5%) and respiratory (15.2%) symptoms. A total of 1534 diagnoses were coded, on average 1.5 per encounter. The predominant diagnostic groups, i.e. ICD-10 chapters, were musculoskeletal (17.2%) and respiratory (12.4%). The most common specific diagnoses were essential hypertension (8.1%) and acute upper respiratory infections (3.7%). A total of 1687 prescriptions were issued, on average 1.6 per encounter. The most frequent pharmaceutical groups were nervous (17.7%), respiratory system (16.2%), and cardiovascular (15.7%). The most frequent drugs were phenoxymethyl penicillin (3.7%), diclofenac (2.9%) and acetylsalicylic acid (2.5%). An average of 1.3 laboratory tests was performed per encounter. In 7.5% of encounters, radiology referrals were made; in 12.3% referrals were made to other specialists/therapists, while sick-list certificates were written in 11.7%. There were significant differences concerning symptoms, diagnoses and investigations between female and male patients, urban and rural practices and SES.

CONCLUSION

The musculoskeletal, respiratory and circulatory systems predominated, both as reasons for the encounter and in the diagnoses, but with significant differences concerning gender, SES and practice location.

Readiness of the Belgian network of sentinel general practitioners to deliver electronic health record data for surveillance purposes: results of survey study

Background

In order to proceed from a paper based registration to a surveillance system that is based on extraction of electronic health records (EHR), knowledge is needed on the number and representativeness of sentinel GPs using a government-certified EHR system and the quality of EHR data for research, expressed in the compliance rate with three criteria: recording of home visits, use of prescription module and diagnostic subject headings.

Methods

Data were collected by annual postal surveys between 2005 and 2009 among all sentinel GPs. We tested relations between four key GP characteristics (age, gender, language community, practice organisation) and use of a certified EHR system by multivariable logistic regression. The relation between EHR software package, GP characteristics and compliance with three quality criteria was equally measured by multivariable logistic regression.

Results

A response rate of 99% was obtained. Of 221 sentinel GPs, 55% participated in the surveillance without interruption from 2005 onwards, i.e. all five years, and 78% were participants in 2009. Sixteen certified EHR systems were used among 91% of the Dutch and 63% of the French speaking sentinel GPs. The EHR software package was strongly related to the community and only one EHR system was used by a comparable number of sentinel GPs in both communities. Overall, the prescription module was always used and home visits were usually recorded. Uniform subject headings were only sometimes used and the compliance with this quality criterion was almost exclusively related to the EHR software package in use.

Conclusions

The challenge is to progress towards a sentinel network of GPs delivering care-based data that are (partly) extracted from well performing EHR systems and still representative for Belgian general practice.

Data extraction from a semi-structured electronic medical record system for outpatients: A model to facilitate the access and use of data for quality control and research

The use of clinical data from electronic medical records (EMRs) for clinical research and for evaluation of quality of care requires an extraction process. Many efforts have failed because the extracted data seemed to be unstructured, incomplete and ridden by errors. We have developed and tested a concept of extracting semi-structured EMRs (Journal III®, Profdoc®) data from 776 diabetes patients in a general practice clinic over a 5 year period. We used standard database management techniques commonly applied in clinical research in the pharmaceutical industry to clean up the data and make the data available for statistical analysis. The key problem was difficulties locating the data, as no standard way to enter the data in the EMR system was reinforced. Furthermore, no built-in edit checks to facilitate data entry were available. Laboratory, drug information and diagnostic data could be used directly while other data such as vital signs required much work to locate and become useful.