Overview of data collection in Bangladesh
| Stage | Method | Target group | Aim |
|---|---|---|---|
| *An asset based poverty score which included literacy, educational attainment, toilet type, and roof type was validated against Demographic Health Survey data. Possible poverty scores ranged from 10 (poorest) to 39 (wealthiest).16 | |||
| †24 hours was used as an indication of severity. A census and expert advice determined that women were unlikely to remain in the hospital beyond 24 hours unless they had a serious complication.16 | |||
| ‡All data collection was undertaken by EP. | |||
| (1) Demographic data collection | Questionnaire (n = 638) | All women admitted to obstetric ward during data collection period | To collect baseline demographic data and poverty related questions in order to calculate a “poverty score”* |
| (2) Consent for follow up | Questionnaire | Women staying>24 hours† | To gain consent for follow up after discharge and get details of address and directions |
| (3) Case note review | Data extraction form (n = 261) | Women staying>24 hours | To examine diagnoses, treatment, delays and deteriorations |
| (4) Follow up after discharge | Semi-structured interviews‡ (n = 25) | Women purposively selected | To explore in more depth the “journey of care” |
| (5) Staff study | Formal (n = 14) and informal interviews‡ | Staff selected from a range of posts and areas relevant to obstetric patients | To explore support mechanisms for those least able to pay |
| (6) Observation | Non-participant observation‡ | To gain a deeper insight into patients’ “journey of care”, routines within the hospital, and support mechanisms available | |