Summary of findings 3. Geriatrician‐led inpatient management compared to orthopaedic‐led inpatient management for adults with dementia following hip fracture surgery.

Geriatrician‐led inpatient management compared to orthopaedic‐led inpatient management for adults with dementia following hip fracture surgery
Patient or population: adults with dementia following hip fracture surgery Setting: hospital ward (inpatient) Intervention: geriatrician‐led inpatient management Comparison: orthopaedic‐led inpatient management
Outcomes	Anticipated absolute effects* (95% CI)		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)	Comments
	Risk with orthopaedic‐led inpatient management	Risk with geriatrician‐led inpatient management
Health‐related quality of life	‐	‐	‐	‐	‐	No data available on this outcome.
Activities of daily living and functional performance (activities of daily living capabilities assessed using BADLS at 12 months) BADLS (0 to 60 points; higher scores equate to poorer functional performance)	The mean BADLS function in the control group was 11.0 points.	The mean BADLS function was 1.5 points lower (3.92 lower to 0.92 points higher).	‐	87 (1 RCT)	⊕⊝⊝⊝ VERY LOW1	The evidence for this result is very uncertain.
Activities of daily living and functional performance (NEADL score at 12 months) NEADL (0 to 22 points, where higher scores equate to greater functional performance)	The mean NEADL function in the control group was 13.6 points.	The mean NEADL function was 3 points lower (8.11 lower to 2.11 points higher).	‐	87 (1 RCT)	⊕⊝⊝⊝ VERY LOW1	The evidence for this result is very uncertain.
Activities of daily living and functional performance (functional performance measured using the SPPB at 12 months) SPPB (0 to 12 points, where higher scores equate to greater functional performance)	The mean SPPB function in the control group was 1.9 points.	The mean SPPB function was 0.3 point higher (0.65 lower to 1.25 points higher).	‐	87 (1 RCT)	⊕⊝⊝⊝ VERY LOW2	The evidence for this result is very uncertain.
Cognitive function (cognitive function measured using the IQCODE at 12 months) IQCODE (1 to 5 points, where higher scores equate to poorer cognitive function)	The mean IQCODE score in the control group was 4.7 points.	The mean cognitive score was 0.1 points higher (0.18 lower to 0.38 higher).	‐	87 (1 RCT)	⊕⊝⊝⊝ VERY LOW1	The evidence for this result is very uncertain.
Behaviour	‐	‐	‐	‐	‐	No data available on this outcome.
Pain	‐	‐	‐	‐	‐	No data available on this outcome.
Mortality at 12 months	33/41 (80%) participants died.	31/46 (67%) participants died.	OR 2.00 (0.74 to 5.36)	87 (1 RCT)	⊕⊝⊝⊝ VERY LOW1	The evidence for this result is very uncertain.
Adverse events (delirium during hospitalisation)	Study population		OR 0.94 (0.52 to 1.72)	212 (2 RCTs)	⊕⊝⊝⊝ VERY LOW3	The evidence for this result is very uncertain.
	721 participants per 1000	708 participants per 1000 (573 to 816)
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). BADLS: basic activities of daily living; CI: confidence interval; IQCODE: Informant Questionnaire on Cognitive Decline in the Elderly; NEADL: Nottingham Extended Activities of Daily Living Scale; OR: odds ratio; RCT: randomised controlled trial; SPPB: Short Performance Physical Battery
GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: We are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: Our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: We have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.

¹Downgraded by three levels from high‐ to very low‐certainty in the outcome due to serious concern about risk of bias (performance, attrition, and contamination bias) and very serious concern about imprecision due to small sample size, data from a single trial, and wide confidence interval (Wyller 2012).
 ²Downgraded by three levels from high‐ to very low‐certainty in the outcome due to serious concern about risk of bias (performance, attrition, and contamination bias) and very serious concern about imprecision due to small sample size, with the data originating from a single trial (Wyller 2012).
 ³Downgraded by three levels from high‐ to very low‐certainty in the outcome due to serious concern about risk of bias (performance and contamination bias (Marcantonio 2001; Wyller 2012), attrition bias (Wyller 2012), and a risk of baseline imbalance (Marcantonio 2001)) and very serious concern about imprecision due to small sample size, data from a single trial, and wide confidence interval (Wyller 2012).