Little is known about the prevalence of depression in people with diabetes in Bangladesh. This study examined the prevalence and factors associated with depression in out-patients with Type 2 diabetes in Bangladesh. In this cross-sectional study a random sample of 483 diabetes out-patients from three diabetes clinics in Bangladesh was invited to participate. Of them 417 patients took part. Depressive symptoms were measured using previously developed and culturally standardized Bengali and Sylheti versions of the World HealthOrganization-5 Well Being Index (WHO-5) and the Patient Health Questionairre-9 (PHQ-9) with predefined cut-off scores. Data was collected using two different modes; e.g. standard assisted collection and audio questionnaire methods. Associations between depression and patient characteristics were explored using regression analysis. The prevalence of depressive symptoms was 34% (PHQ-9 score ≥ 5) and 36% (WHO-5 score < 52) with audio questionnaire delivery method. The prevalence rates were similar regardless of the type (PHQ-9 vs. WHO-5) and language (Sylheti vs. Bengali) of the questionnaires, and methods of delivery (standard assisted vs. audio methods). The significant predictors of depressive symptoms using either the PHQ-9 or WHO-5 questionnaires were; age, income, gender, treatment intensity, and co-morbid cardiovascular disease. Further, depression was strongly associated with poor glycaemic control and number of co-morbid conditions. This study demonstrated that depression prevalence is common in out-patients with type 2 diabetes in Bangladesh. In a setting where recognition, screening and treatment levels remain low, health care providers need to focus their efforts on diagnosing, referring and effectively treating this important disease in order to improve service delivery.

Background: Previous health research has often explicitly excluded individuals from minority ethnic backgrounds due to perceived cultural and communication difficulties, including studies where there might be language/literacy problems in obtaining informed consent. This study addressed these difficulties by developing audio-recorded methods of obtaining informed consent and recording data. This report outlines 1) our experiences with securing recruitment to a qualitative study investigating alternative methods of data collection, and 2) the development of a standardised process for obtaining informed consent from individuals from minority ethnic backgrounds whose main language does not have an agreed written form. Methods: Two researchers from South Asian backgrounds recruited adults with Type 2 diabetes whose main language was spoken and not written, to attend a series of focus groups. A screening tool was used at recruitment in order to assess literacy skills in potential participants. Informed consent was obtained using audio-recordings of the patient information and recording patients' verbal consent. Participants' perceptions of this method of obtaining consent were recorded. Results: Recruitment rates were improved by using telephone compared to face-to-face methods. The screening tool was found to be acceptable by all potential participants. Audio-recorded methods of obtaining informed consent were easy to implement and accepted by all participants. Attrition rates differed according to ethnic group. Snowballing techniques only partly improved participation rates. Conclusion: Audio-recorded methods of obtaining informed consent are an acceptable alternative to written consent in study populations where literacy skills are variable. Further exploration of issues relating to attrition is required, and a range of methods may be necessary in order to maximise response and participation rates.

Background: Understanding injection practices is crucial for evidence-based development of intervention initiatives. This study explored the extent of injection use and injection safety practices in primary care hospitals in Bangladesh. Methods: The study employed both quantitative and qualitative research methods. The methods used were - a retrospective audit of prescriptions (n = 4320), focus group discussions (six with 43 participants), in-depth interviews (n = 38) with a range service providers, and systematic observation of the activities of injection providers (n = 120), waste handlers (n = 48) and hospital facilities (n = 24). Quantitative and qualitative data were assessed with statistical and thematic analysis, respectively, and then combined. Results: As many as 78% of our study sample (n = 4230) received an injection. The most commonly prescribed injections (n = 3354) including antibiotics (78.3%), IV fluids (38.6%), analgesics/pain killers (29.4%), vitamins (26.7%), and anti-histamines (18.5%). Further, 43.7% (n = 1145) of the prescribed antibiotics (n = 2626) were given to treat diarrhea and 42.3% (n = 600) of IV fluids (n = 1295) were used to manage general weakness conditions. Nearly one-third (29.8%; n = 36/120) of injection providers reported needle-stick injuries in the last 6 months with highest incidences in Rajshahi division followed by Dhaka division. Disposal of injection needles, syringes and other materials was not done properly in 83.5% (n = 20/24) of the facilities. Health providers' safety concerns were not addressed properly; only 23% (n = 28/120) of the health providers and 4.2% (n = 2/48) of the waste handlers were fully immunized against Hepatitis B virus. Moreover, 73% (n = 87/120) of the injection providers and 90% (n = 43/48) of the waste handlers were not trained in injection safety practices and infection prevention. Qualitative data further confirmed that both providers and patients preferred injections, believing that they provide quick relief. The doctors' perceived injection use as their prescribing norm that enabled them to prove their professional credibility and to remain popular in a competitive health care market. Additionally, persistent pressure from hospital administration to use up injections before their expiry dates also influenced doctors to prescribe injections regardless of actual indications. Conclusions: As far as the patients and providers' safety is concerned, this study demonstrated a need for further research exploring the dynamics of injection use and safety in Bangladesh. In a context where a high level of injection use and unsafe practices were reported, immediate prevention initiatives need to be operated through continued intervention efforts and health providers' training in primary care hospitals in Bangladesh.