Hopkins 2008 UGA (Medium).
| Methods | Trial design: cluster randomized pre‐post open‐label trial; data included in this review are from the two months following introduction of RDTs in the intervention arm. | |
| Participants | Number of participants randomized: total fever cases 1550 (602 in RDT arm, 948 in presumptive treatment arm) Number of participants randomized for clinical outcomes was 25% of total fever cases, i.e.388 (151 versus 267 respectively) Number of participants analysed for primary outcomes: (a) prescribing of antimalarials analysis 1550 (0% loss); (b) clinical outcomes: 328 (15.4% loss: 6.3% in intervention arm versus 21.1% in presumptive treatment arm) Inclusion: Any patient deemed eligible for RDT testing by the healthworker Exclusion: None |
|
| Interventions | Intervention: Training in fever case management based on RDTs. Control: Standard‐of‐care symptom‐based or empiric treatment of fever. The intervention group received training and RDTs; one‐day follow‐up support supervision was conducted two weeks after the initial three‐day training. Data collection commenced after the follow‐up support supervision visit. The control group continued usual symptom‐based care according to existing Uganda Ministry of Health guidelines. RDT performed by treating clinician (usually clinical officer or nursing staff). No other formal supervision was provided. |
|
| Outcomes | Primary: patients with fever on day 4 Secondary:
|
|
| Notes | Country: Uganda RDT: paracheck (HRP2) Setting: peripheral health centres Sampling: Lack of microscopy services, at least 3 full‐time clinical staff, estimated patient volume of at least 200 patients per week, willingness of health centre staff to participate in the trial, and location within 20 km of a sentinel health centre established by the Uganda Malaria Surveillance Project. Transmission: Medium; reference slide positivity in all participants with fever 32% Dates: 2008; first half Funding: Exxon Mobil Corp. via the Academic Alliance Foundation; and NIH, USA, K23 AI065457‐01 |
|
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Low risk | Coin flip in the presence of health centre leaders. |
| Allocation concealment (selection bias) | Unclear risk | Allocation of RDTs was decided by coin flip in the presence of study staff and representatives from each matched pair of health centers |
| Blinding of participants and personnel (performance bias) All outcomes | High risk | Both the trial participants and personnel were aware of intervention allocation |
| Blinding of outcome assessment (detection bias) All outcomes | High risk | Both the trial participants and personnel were aware of the diagnosis made and treatment prescribed. |
| Incomplete outcome data (attrition bias) All outcomes | Low risk | Loss to follow‐up was low (0.0% for prescribing of antimalarials; 15.4% for clinical outcomes, differentiated by trial group) Performed available case analysis, |
| Selective reporting (reporting bias) | Low risk | Reported on all trial outcomes described in the methodology |
| Other bias | Low risk | No other sources of bias identified |