Pitkala 2014.
| Methods | Cluster‐RCT (randomised by ward) Total duration of study: 12 months |
|
| Participants | 227 residents in 20 facilities (10 control, 10 intervention) Setting: Assisted living facilities Age: Control mean 83.5 SD 6.9; intervention mean 82.9 years SD 7.5 Gender: Control 77.1% female; intervention 65.3% female Country: Finland Date of Study: Not stated |
|
| Interventions | Educational intervention: Two 4‐hour training sessions for nursing staff. Aim of session was to enable nurses to recognise harmful medications and corresponding adverse drug events. First 4‐hour session: lecture‐based, allowed participants to discuss medication‐related problems experienced by their own residents, introduced lists of harmful medications and suitable treatments. Also involved discussion about medication use for residents with real impairment and drug‐drug interactions Second 4‐hour session: case‐study‐based, demonstrate relevance and importance of topic to nurses During both training sessions nurses were encouraged to reflect on their own procedure and opportunities for improvement Those nurses that received this intervention were asked to identify potential medication‐related problem and highlight these to the consulting physician |
|
| Outcomes | Assessed at 0, 6, 12 months Hospital admissions (hospital days) Mortality Health‐related Quality of Life (15D) Medication appropriateness (composite of Beers criteria, Anticholinergic Risk Scale, > 2 psychotropic medications, NSAIDs and proton pump inhibitors) Not used in this review: Cognitive assessment (MMSE) Nutritional assessment (Mini‐nutritional assessment) |
|
| Notes | [author contacted] | |
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Low risk | Computerised random number generator |
| Allocation concealment (selection bias) | Low risk | Cluster design |
| Blinding of participants and personnel (performance bias) All outcomes | High risk | Blinding not conducted |
| Blinding of outcome assessment (detection bias) Subjective outcomes | Unclear risk | Insufficient information to permit judgement |
| Blinding of outcome assessment (detection bias) Objective outcomes | Low risk | Outcomes not likely to be influenced by lack of blinding |
| Incomplete outcome data (attrition bias) Primary outcomes | Unclear risk | Reasons and proportions for attrition documented and similar in intervention and control. Described as intention‐to‐treat analysis by authors Overall attrition rate relatively high (27.8%) |
| Incomplete outcome data (attrition bias) Secondary outcomes | Unclear risk | Reasons and proportions for attrition documented and similar in intervention and control. Described as modified intention‐to‐treat analysis by authors Overall attrition rate relatively high (27.8%) |
| Selective reporting (reporting bias) | Unclear risk | Insufficient information to permit judgement |
| Similar baseline outcome measurements | Unclear risk | Lower HRQoL in intervention group (15D score mean 0.61 [SD 0.12] vs 0.66 [0.11]) and higher mean number of harmful medications (2.9 [SD 1.8] vs 2.5 [SD 1.7]). Analyses were adjusted for these differences |
| Similar baseline characteristics | Unclear risk | More males (34.7% vs 22.9%), higher prevalence of ‘as‐needed’ medication (mean 3.6 [SD 2.3] vs 2.9 [SD2.0]), and higher number of comorbidities (Mean Charlson’s index 3.2 [2.0] vs 2.5 [1.8]) in intervention group. Analyses were adjusted for these differences |
| Reliable primary outcome measure | Low risk | Well‐defined potentially harmful medication use |
| Adequate protection against contamination | Unclear risk | Cluster design. However, it was theoretically possible that some nurses may have moved between intervention and control nursing homes, although this was deemed unlikely by the author [author contacted] |
| Other bias | Low risk | Appears to be free from other sources of bias |