Describing diversity of real world data sources in pharmacoepidemiologic studies: The DIVERSE scoping review.

PURPOSE: Real-world evidence (RWE) is increasingly used for medical regulatory decisions, yet concerns persist regarding its reproducibility and hence validity. This study addresses reproducibility challenges associated with diversity across real-world data sources (RWDS) repurposed for secondary use in pharmacoepidemiologic studies. Our aims were to identify, describe and characterize practices, recommendations and tools for collecting and reporting diversity across RWDSs, and explore how leveraging diversity could improve the quality of evidence. METHODS: In a preliminary phase, keywords for a literature search and selection tool were designed using a set of documents considered to be key by the coauthors. Next, a systematic search was conducted up to December 2021. The resulting documents were screened based on titles and abstracts, then based on full texts using the selection tool. Selected documents were reviewed to extract information on topics related to collecting and reporting RWDS diversity. A content analysis of the topics identified explicit and latent themes. RESULTS: Across the 91 selected documents, 12 topics were identified: 9 dimensions used to describe RWDS (organization accessing the data source, data originator, prompt, inclusion of population, content, data dictionary, time span, healthcare system and culture, and data quality), tools to summarize such dimensions, challenges, and opportunities arising from diversity. Thirty-six themes were identified within the dimensions. Opportunities arising from data diversity included multiple imputation and standardization. CONCLUSIONS: The dimensions identified across a large number of publications lay the foundation for formal guidance on reporting diversity of data sources to facilitate interpretation and enhance replicability and validity of RWE.

Outcome misclassification: Impact, usual practice in pharmacoepidemiology database studies and an online aid to correct biased estimates of risk ratio or cumulative incidence.

PURPOSE: It is well documented that outcome misclassification can bias a point estimate. We aimed to understand current practice in addressing this bias in pharmacoepidemiology database studies and to develop an open source application (app) from existing methodology to demonstrate the impact and mechanism of this bias on results. METHODS: Studies of an exposure and a clinical outcome were selected from all Pharmacoepidemiology and Drug Safety publications during 2017 and any reference to outcome misclassification described. An app to correct risk ratio (RR) and cumulative incidence for outcome misclassification was developed from a published methodology and used to demonstrate the impact of correction on point estimates. RESULTS: Eight (19%) of 43 papers selected reported estimates of outcome ascertainment accuracy with positive predictive value (PPV) the most commonly reported measure (7 of 8 studies). Three studies (7%) corrected for the bias, 1 by exposure strata, and 5 (12%) restricted analyses to confirmed cases. The app (app http://apps.p-95.com/ISPE/) uses values of PPV and sensitivity (or a range of possible values) in each exposure strata and returns corrected point estimates and confidence intervals. The app demonstrates that small differences between comparison groups in PPV or sensitivity can introduce bias even when accuracy estimates are high. CONCLUSIONS: Outcome misclassification is not usually corrected in pharmacoepidemiology database studies although correction methods using routinely measured indices are available. Error indices are needed for each comparison group to correct RR estimates for these errors. The app should encourage understanding of this bias and increase adjustment.

Incorporating patient generated health data into pharmacoepidemiological research.

Epidemiology and pharmacoepidemiology frequently employ Real-World Data (RWD) from healthcare teams to inform research. These data sources usually include signs, symptoms, tests, and treatments, but may lack important information such as the patient's diet or adherence or quality of life. By harnessing digital tools a new fount of evidence, Patient (or Citizen/Person) Generated Health Data (PGHD), is becoming more readily available. This review focusses on the advantages and considerations in using PGHD for pharmacoepidemiological research. New and corroborative types of data can be collected directly from patients using digital devices, both passively and actively. Practical issues such as patient engagement, data linking, validation, and analysis are among important considerations in the use of PGHD. In our ever increasingly patient-centric world, PGHD incorporated into more traditional Real-Word data sources offers innovative opportunities to expand our understanding of the complex factors involved in health and the safety and effectiveness of disease treatments. Pharmacoepidemiologists have a unique role in realizing the potential of PGHD by ensuring that robust methodology, governance, and analytical techniques underpin its use to generate meaningful research results.

Evaluation of a new patient consultation initiative in community pharmacy for ear, nose and throat and eye conditions.

BACKGROUND: Community pharmacy Common Ailments Services can ease the considerable workload pressures on primary and secondary care services. However, evidence is needed to determine whether there are benefits of extending such services beyond their typically limited scope. This study therefore aimed to evaluate a new community pharmacy model of a service for patients with ear, nose and throat (ENT) and eye conditions who would otherwise have had to seek primary care appointments or emergency care. METHODS: People with specified ENT or eye conditions registered with General Practitioners in Staffordshire or Shropshire who presented at participating community pharmacies were offered a consultation with a pharmacist trained to provide the service. The service included provision of relevant self-care advice and, where clinically appropriate, supply of non-prescription medicines or specified prescription-only medicines (POMs), including antibiotics, under Patient Group Directions. Patients received a follow up telephone call from the pharmacist five days later. Data were collected on the characteristics of patients accessing the service, the proportion of those who were treated by the pharmacist without subsequently seeing another health professional about the same condition, and patient reported satisfaction from a questionnaire survey. RESULTS: A total of 408 patients accessed the service, of whom 61% received a POM, 15% received advice and medicine supplied under the common ailments service, 9% received advice and purchased a medicine, 10% received advice only and 5% were referred onwards. Sore throat accounted for 45% of diagnoses where a POM was supplied, 32% were diagnosed with acute otitis media and 15% were diagnosed with acute bacterial conjunctivitis. The number of patients successfully followed up was 309 (76%), of whom 264 (85%) had not seen another health professional for the same symptoms, whilst 45 (15%) had seen another health professional, usually their GP. The questionnaire was completed by 259 patients (response rate 63%) of whom 96% reported being very satisfied or satisfied with the service. CONCLUSIONS: The study demonstrates that pharmacists can effectively diagnose and treat these conditions, with a high degree of patient satisfaction. Wider adoption of such service models could substantially benefit primary care and emergency care services.

Outcomes of salvage surgery for the oropharynx and larynx: a contemporary experience in a UK Cancer Centre.

INTRODUCTION: The purpose of this study was to review our recent experience of salvage surgery, comparing larynx and oropharynx recurrence patterns. METHODS: A single centre, retrospective review of salvage surgery for recurrent head and neck cancer including patients between 2008 and 2016. RESULTS: 61 patients were identified, 36 underwent salvage laryngectomy and 25 received oropharyngeal resections. The median overall survival of oropharyngeal recurrent tumors was 26 months (95% CI 15-118 months) and for laryngeal tumors was 23 months (95% CI 11-38 months), p = 0.1008. There was a significant overall survival benefit in patients with negative resection margin. The median survival in the negative margin group was 38 months (95% CI 25-108 months) compared to the positive margin group, 9 months (95% CI 5-15 months), p < 0.0001. CONCLUSION: Survival results following surgical salvage in the larynx and oropharynx appear to be similarly poor. Those patients with clear margins appear to have a significantly better prognosis.

Observational safety study of specific outcomes after trivalent cell culture seasonal influenza vaccination (Optaflu® ) among adults in THIN database of electronic UK primary healthcare records.

PURPOSE: To investigate the safety of trivalent seasonal influenza vaccine (TIVc) (Optaflu® ), the first cell culture seasonal trivalent influenza vaccine available in Europe. METHODS: Codes and unstructured text in adult electronic healthcare records (The Health Improvement Network) were searched for a TIVc brand name or batch number and possible outcomes within a 3 month pre- to 6 month post-TIVc exposure study period (2012-2015). The outcomes were severe allergic reactions, Bell's palsy, convulsions, demyelination, paresthesia, noninfectious encephalitis, neuritis (optic and brachial), vasculitis, inflammatory bowel disease, and thrombocytopenia. Risk periods were defined based on biologically plausible time frame postvaccination when an outcome caused by the vaccine might be expected to occur. Possible outcomes were adjudicated against outcome specific case definitions and a date of onset assigned by using electronic and other medical records. Observed (risk period) to expected (outside risk and preexposure periods) rate ratios, postexposure incidence, and plots of time from exposure to outcome were reported. RESULTS: Sixteen of 1011 events from 4578 exposures fulfilled a primary case definition and had a date of onset during the study period. Three were in observed time. The observed-to-expected rate ratios were (3.3, 95% CI 0.3, 31.7) for convulsions and (1.5, 95% CI 0.2, 14.9) for thrombocytopenia with 1 outcome each in observed time. There was 1 incident inflammatory bowel disease in observed, but none in expected, time. CONCLUSION: The small sample size restricts interpretation; however, no hypothesis of an increased risk of a study outcome was generated. Adjudication of events against case definitions to reduce misclassification of onset and outcomes allowed use of precise risk periods. KEY POINTS This observational study did not generate a hypothesis of an association between the first cell-culture seasonal influenza vaccination available in the European Union and any of the study outcomes (severe allergic reactions, Bell's palsy, convulsions, demyelination, paresthesia, noninfectious encephalitis, neuritis [optic and brachial], vasculitis, inflammatory bowel disease [IBD], and thrombocytopenia). The small sample size limits interpretation of the results. The review of each possible outcome identified from electronic healthcare records against case definitions was included to minimize misclassification of time and outcomes and allow the use of precise risk-periods in an observed-to-expected within cohort analysis. Plots of time from exposure to outcome were included to assess the risk windows.

The incidence of Kawasaki disease after vaccination within the UK pre-school National Immunisation Programme: an observational THIN database study.

PURPOSE: To provide expected incidence rates of Kawasaki disease after vaccination in routine clinical practice and as recommended within a pre-school National Immunisation Programme (NIP). METHODS: A post-immunisation risk period when Kawasaki disease onset might be associated with vaccination was defined as 28 days. Immunisation records for children under 6 years were identified from The Health Improvement Network (THIN) database of electronic UK primary health care records (2008-2012) and linked to previously validated cases of Kawasaki disease with an assigned date of onset. Kawasaki disease incidence in the risk period after a complete NIP recommended set of vaccinations was estimated for five vaccination stages individually and in total. RESULTS: A total of 642 170 complete pre-school immunisation stages from 275 986 children were included. Six cases of Kawasaki disease had onset in the risk period after any NIP stage providing an incidence of 12.8 per 100 000 person years (95%CI 5.7, 28.4). The incidence after any single immunisation stage ranged from 0 to 27.4 (95%CI 8.8, 84.8) per 100 000 person years. CONCLUSION: There were few cases of Kawasaki disease in the risk period after any NIP vaccination combination. The incidence rates will aid in the interpretation of clinical trials and post-marketing surveillance of new vaccines. Copyright © 2016 John Wiley & Sons, Ltd.

Evidence generation from healthcare databases: recommendations for managing change.

There is an increasing reliance on databases of healthcare records for pharmacoepidemiology and other medical research, and such resources are often accessed over a long period of time so it is vital to consider the impact of changes in data, access methodology and the environment. The authors discuss change in communication and management, and provide a checklist of issues to consider for both database providers and users. The scope of the paper is database research, and changes are considered in relation to the three main components of database research: the data content itself, how it is accessed, and the support and tools needed to use the database. Copyright © 2016 John Wiley & Sons, Ltd.

Foodborne illness, Australia, circa 2000 and circa 2010.

Foodborne disease is a major public health problem worldwide. To examine changes in foodborne illness in Australia, we estimated the incidence, hospitalizations, and deaths attributed to contaminated food circa 2010 and recalculated estimates from circa 2000. Approximately 25% of gastroenteritis cases were caused by contaminated food; to account for uncertainty we used simulation techniques to estimate 90% credible intervals. We estimate that circa 2010, 4.1 million foodborne gastroenteritis cases occurred, and circa 2000, 4.3 million cases occurred. Circa 2010, contaminated food was estimated to be responsible for 30,840 gastroenteritis-associated hospitalizations, 76 associated deaths, and 5,140 nongastrointestinal illnesses. Cases of salmonellosis and campylobacteriosis increased from 2000 to 2010 and were the leading causes of gastroenteritis-associated hospitalizations; Listeria monocytogenes and nontyphoidal Salmonella spp. infections were the leading causes of death. Although the overall incidence of foodborne illnesses declined over time in Australia, cases of foodborne gastroenteritis are still common.

Sequelae of foodborne illness caused by 5 pathogens, Australia, circa 2010.

In Australia circa 2010, 4.1 million (90% credible interval [CrI] 2.3-6.4 million) episodes of foodborne gastroenteritis occurred, many of which might have resulted in sequelae. We estimated the number of illnesses, hospitalizations, and deaths from Guillain-Barré syndrome, hemolytic uremic syndrome, irritable bowel syndrome, and reactive arthritis that were associated with contaminated food in Australia. Data from published studies, hospital records, and mortality reports were combined with multipliers to adjust for different transmission routes. We used Monte Carlo simulation to estimate median estimates and 90% CrIs. In Australia, circa 2010, we estimated that 35,840 (90% CrI 25,000-54,000) illnesses, 1,080 (90% CrI 700-1,600) hospitalizations, and 10 (90% CrI 5-14) deaths occurred from foodborne gastroenteritis-associated sequelae. Campylobacter spp. infection was responsible for 80% of incident cases. Reducing the incidence of campylobacteriosis and other foodborne diseases would minimize the health effects of sequelae.

Human papillomavirus infection is rare in nonmalignant tonsil tissue in the UK: implications for tonsil cancer precursor lesions.

The incidence of human papillomavirus (HPV)-associated tonsil cancer is increasing but the prevalence of HPV, and of premalignant precursors, in tonsil tissue is unknown. We aimed to assess prevalence of HPV infection in nonmalignant tonsillar crypt epithelia and to histopathologically characterise positive samples. Formalin-fixed paraffin-embedded (FFPE) tonsil tissue specimens were obtained from an age- and sex-stratified random sample of patients aged 0-69 years whose paired tonsils were archived following elective tonsillectomy at hospitals throughout England and Southern Scotland from 2004 to 2008. Homogenised fresh-frozen tonsil tissue was also obtained from archive for two random subsets of males aged 25-34 and over 44. HPV status was assessed in all samples for 20 mucosal HPV types by GP5+/6+ polymerase chain reaction (PCR) enzyme immunoassay and by HPV16 type-specific PCR targeting the E6 gene. In the homogenised material, HPV status was also assessed for 44 HPV types by SPF10-PCR enzyme immunoassay. Of 4,095 randomly sampled FFPE specimens, amplifiable DNA was extracted from 3,377 (82.5%) and from 511 of 524 (97.5%) homogenised tonsils. HPV DNA was identified in 0 of 3,377 (0%, 95% CI 0-0.089%) fixed samples and 0 of 511 (0%, 95% CI 0-0.58%) homogenised samples. This suggests HPV infection may be rare in tonsil reticulated crypt epithelia. Furthermore, we found no evidence of HPV-associated premalignant neoplasia. These data suggest that if HPV-associated premalignant lesions do occur, they are likely to be rare and may have a high risk of progression to carcinoma.

Descriptive investigation of the recording of influenza vaccination details on The Health Information Network database.

BACKGROUND: The UK primary care databases are used in pharmacoepidemiology studies of vaccination type. We investigated vaccine recording and whether, and how, exposure to specific brands and batches can be identified. METHODS: Details of influenza vaccinations given in the 2010-2011 or 2011-2012 seasons were identified from coded and text fields in The Health Information Network UK primary care database. The proportion of people over 64 years of age vaccinated against influenza was compared with published regional rates. Searches for Fluvirin (Novartis Vaccines and Diagnostics GmbH, Marburg, Germany) batch numbers and name identified exposure to this specific vaccine. The recording of any brand name and batch number was described for a sample of 1000 vaccinations across 472 practices. RESULTS: A total of 767 904 influenza vaccinations were identified during the 2010-2011 season and 784 518 in 2011-2012. Vaccination rates for people aged over 64 years were 75.6%, 80.9%, 78.4% and 71.9% in England, Northern Ireland, Scotland and Wales, respectively (2011-2012 season), compared with published figures of 74.0%, 77.0%, 76.2% and 67.7%. Rates were slightly lower in 2010-2011 in both data sources. A Fluvirin brand was identified for 3.6% of all UK vaccinations but 26.2% of those in Scottish practices. Vaccination brand could be identified for 94.3% of the sample, 93.6% with a batch number. Batch number (98.5%) and brand name (50.3%) were most frequently recorded in an immunisation 'batch' text field. CONCLUSION: Patients exposed to an influenza vaccine in primary care can be identified from The Health Information Network. Identification of brand or batch number requires a text search. Regional variation in brand of vaccine should be considered when estimating sample size.

Proportion of illness acquired by foodborne transmission for nine enteric pathogens in Australia: an expert elicitation.

BACKGROUND: Estimates of the burden of illness acquired from food inform public health policy and prioritize interventions. A key component of such estimates is the proportion of illnesses that are acquired by foodborne transmission. In view of the shortage of requisite data, these proportions are commonly obtained through a process known as expert elicitation. We report findings from an elicitation process used to assess the importance of the foodborne transmission route for nine pathogens in Australia, circa 2010. MATERIALS AND METHODS: Eleven experts were asked to estimate the proportion of illness acquired by five transmission routes: food, environmental, water, person, and zoonotic, together with a 90% certainty interval for foodborne transmission. Foodborne estimates and intervals from each expert were combined using both modified triangular and Program Evaluation and Review Technique (PERT) distributions, in @Risk version 6, to generate final distributions from which median estimates and 95% Credible Intervals (CrI) were calculated. RESULTS: Shiga toxin-producing Escherichia coli (STEC) was the only pathogen believed to have an important zoonotic transmission route, while norovirus, hepatitis A virus, non-STEC pathogenic E. coli, and Shigella spp. were all thought to be primarily spread from person to person. Foodborne transmission was the main route for Clostridium perfringens (98%, CrI: 84-100), Listeria monocytogenes (98%, CrI: 86-100), nontyphoidal Salmonella spp. (72%, CrI: 50-87), and Campylobacter spp. (77%, CrI: 60-90). Foodborne estimates using the modified triangular distribution had wider CrI than these calculated using the PERT distribution. CONCLUSIONS: Foodborne proportions for most pathogens in this study were the same or lower than those estimated circa 2000 in Australia, with the greatest decline for non-STEC pathogenic E. coli. Inclusion of certainty intervals from experts helps to quantify the precision of foodborne proportions. A decline in estimates of the foodborne proportion for common pathogens will influence final estimates of the burden of illness acquired from food.

Milk supply related to childhood leukaemia treatment.

This literature review and case study answers the question: 'Do the late effects of childhood cranial radiation therapy include impacts on breastfeeding?' PubMed was searched for papers using the terms lactation and cranial radiotherapy or childhood cranial radiotherapy. The case study was written from one author's experience of helping a mother with a history of childhood cranial radiation therapy. The few available studies report a high rate of lactation failure in women who were treated with cranial radiation therapy for childhood cancer, but the exceptions indicate that lactation failure is not inevitable in this group of mothers. Breastfeeding may ameliorate some of the adverse effects of cranial radiation therapy. Health professionals caring for mothers with a history of cranial radiation therapy must balance encouraging women to breastfeed with preparing them for the possibility that they may be unable to do so.

Developing and Validating a Multivariable Prognostic-Predictive Classifier for Treatment Escalation of Oropharyngeal Squamous Cell Carcinoma: The PREDICTR-OPC Study.

PURPOSE: While there are several prognostic classifiers, to date, there are no validated predictive models that inform treatment selection for oropharyngeal squamous cell carcinoma (OPSCC).Our aim was to develop clinical and/or biomarker predictive models for patient outcome and treatment escalation for OPSCC. EXPERIMENTAL DESIGN: We retrospectively collated clinical data and samples from a consecutive cohort of OPSCC cases treated with curative intent at ten secondary care centers in United Kingdom and Poland between 1999 and 2012. We constructed tissue microarrays, which were stained and scored for 10 biomarkers. We then undertook multivariable regression of eight clinical parameters and 10 biomarkers on a development cohort of 600 patients. Models were validated on an independent, retrospectively collected, 385-patient cohort. RESULTS: A total of 985 subjects (median follow-up 5.03 years, range: 4.73-5.21 years) were included. The final biomarker classifier, comprising p16 and survivin immunohistochemistry, high-risk human papillomavirus (HPV) DNA in situ hybridization, and tumor-infiltrating lymphocytes, predicted benefit from combined surgery + adjuvant chemo/radiotherapy over primary chemoradiotherapy in the high-risk group [3-year overall survival (OS) 63.1% vs. 41.1%, respectively, HR = 0.32; 95% confidence interval (CI), 0.16-0.65; P = 0.002], but not in the low-risk group (HR = 0.4; 95% CI, 0.14-1.24; P = 0.114). On further adjustment by propensity scores, the adjusted HR in the high-risk group was 0.34, 95% CI = 0.17-0.67, P = 0.002, and in the low-risk group HR was 0.5, 95% CI = 0.1-2.38, P = 0.384. The concordance index was 0.73. CONCLUSIONS: We have developed a prognostic classifier, which also appears to demonstrate moderate predictive ability. External validation in a prospective setting is now underway to confirm this and prepare for clinical adoption.

An observational descriptive study of the epidemiology and treatment of neuropathic pain in a UK general population.

BACKGROUND: This study updated our knowledge of UK primary care neuropathic pain incidence rates and prescribing practices. METHODS: Patients with a first diagnosis of post-herpetic neuralgia (PHN), painful diabetic neuropathy (PDN) or phantom limb pain (PLP) were identified from the General Practice Research Database (2006 - 2010) and incidence rates were calculated. Prescription records were searched for pain treatments from diagnosis of these conditions and the duration and daily dose estimated for first-line and subsequent treatment regimens. Recording of neuropathic back and post-operative pain was investigated. RESULTS: The study included 5,920 patients with PHN, 5,340 with PDN, and 185 with PLP. The incidence per 10,000 person-years was 3.4 (95% CI 3.4, 3.5) for PHN; and 0.11 (95% CI 0.09, 0.12) for PLP. Validation of the PDN case definition suggested that was not sensitive. Incident PHN increased over the study period. The most common first-line treatments were amitriptyline or gabapentin in the PDN and PLP cohorts, and amitriptyline or co-codamol (codeine-paracetamol) in PHN. Paracetamol, co-dydramol (paracetamol-dihydrocodeine) and capsaicin were also often prescribed in one or more condition. Most first-line treatments comprised only one therapeutic class. Use of antiepileptics licensed for neuropathic pain treatment had increased since 2002-2005. Amitriptyline was the only antidepressant prescribed commonly as a first-line treatment. CONCLUSION: The UK incidence of diagnosed PHN has increased with the incidence of back-pain and post-operative pain unclear. While use of licensed antiepileptics increased, prescribing of therapy with little evidence of efficacy in neuropathic pain is still common and consequently treatment was often not in-line with current guidance.

Independent reporting in oral and maxillofacial pathology.

AIMS: To understand the current practice, extent of use and barriers related to independent reporting (IR) in oral and maxillofacial pathology (OMFP) training in the UK. METHODS: A questionnaire was created containing questions about the experiences and opinions surrounding IR in OMFP. The target participants were (1) consultants in OMFP who had been involved in training OMFP trainees in the last 5 years and (2) current OMFP trainees. The questionnaire was delivered via Google Forms and disseminated using a link in an invitation email sent to the participants. RESULTS: A total of 13 consultant responses (response rate of 81%) and 12 trainee responses (response rate of 92%) were received. Of these, three consultants and five trainees were using IR at the time of the study. Several themes emerged highlighting the perceived benefits and concerns regarding IR. CONCLUSIONS: This study suggests that there is a disparity in the way IR is used in OMFP training across the UK. There was shared concern between consultants and trainees regarding the lack of clear guidance and subsequent fear of litigation. These are issues that need to be addressed if trainees are to have a similar experience across the country and be prepared for independent practice on completion of training.

Foodborne and waterborne infections in elderly community and long-term care facility residents,Victoria, Australia.

We calculated rates of foodborne and waterborne infections reported to the health department in Victoria, Australia, during 2000-2009 for elderly residents of long-term care facilities (LTCFs) and the community. We used negative binomial regression to estimate incidence rate ratios, adjusting for age, sex, and reporting period. We analyzed 8,277 infections in elderly persons. Rates of campylobacteriosis, legionellosis, listeriosis, toxigenic Escherichia coli infections, and shigellosis were higher in community residents, and rates of Salmonella infection were higher in LTCF residents. Each year, 61.7 Campylobacter infections were reported per 100,000 LTCF residents, compared with 97.6 per 100,000 community residents. LTCF residents were at higher risk for S. enterica serotype Typhimurium associated with outbreaks. Rates of foodborne infections (except salmonellosis) were similar to or lower for LTCF residents than for community residents. These findings may indicate that food preparation practices in LTCFs are safer than those used by elderly persons in the community.

Macrovascular and microvascular outcomes after beginning of insulin versus additional oral glucose-lowering therapy in people with type 2 diabetes: an observational study.

PURPOSE: In type 2 diabetes, the optimal stage to introduce insulin can be unclear. We compared the incidence of subsequent vascular disease between treatment regimens, that is, adding another oral glucose-lowering drug (OGLD) versus starting insulin treatment. METHODS: People with poor control on OGLDs who intensified treatment (2000-2007) were grouped by number of baseline OGLDs. Two composite endpoints, of macrovascular disease (all-cause mortality, myocardial infarction, acute coronary syndrome and stroke) and of microvascular disease (peripheral neuropathy, nephropathy or retinopathy), together with HbA(1c) and weight change over a year, were compared in those beginning insulin versus an additional OGLD. All data came from The Health Information Network UK primary care database. RESULTS: After exclusions, 14,904 people intensified treatment from one OGLD, 7231 from two and 978 from three, 9, 41 and 90%, respectively, started insulin. Average follow-up was 3.5 years. The adjusted hazard ratios for macrovascular events, OGLD versus insulin, were 0.53 (95%CI 0.42, 0.69) from one baseline treatment, 0.85 (0.70 1.04) from two and 1.07 (0.50, 2.30) from three, with no difference in risk of microvascular disease in any comparison. Mean body weight increased, and mean HbA(1c) fell across groups; the only significant adjusted comparison was greater weight increase when commencing insulin from one OGLD. CONCLUSIONS: Starting insulin rather than adding another OGLD to double or triple oral therapy did not significantly increase the incidence of vascular events. Beginning insulin from one OGLD was uncommon. More incident macrovascular disease in this group may be caused by residual confounding.

Guidelines for good database selection and use in pharmacoepidemiology research.

The use of healthcare databases in research provides advantages such as increased speed, lower costs and limitation of some biases. However, database research has its own challenges as studies must be performed within the limitations of resources, which often are the product of complex healthcare systems. The primary purpose of this document is to assist in the selection and use of data resources in pharmacoepidemiology, highlighting potential limitations and recommending tested procedures. This guidance is presented as a detailed text with a checklist for quick reference and covers six areas: selection of a database, use of multiple data resources, extraction and analysis of the study population, privacy and security, quality and validation procedures and documentation.