Table 9.

Multi family therapy for eating disorders

Certainty assessment							Impact	Certainty	Importance
DE	Study design	Risk of bias	Inconsistency	Indirectness	Imprecision	Other considerations
Good Outcome at End of Treatment (assessed with: Morgan Russell Scale), Psychological Symptoms (EDE)
1	randomised trials	not serious	not serious	not serious	not serious	none	RCT (n = 169) of adolescents with AN aged 11–18 comparing MFT to FBT (91% female). 65/86 (75.6%) good outcome at end of treatment in MFT versus 48/83 (57.8%) in the FBT group - significant difference. No differences between groups seen on the EDE. Both groups improved over time on the EDE.	⨁⨁⨁⨁ HIGH	CRITICAL
		not serious	not serious	not serious	not serious	none	No differences between groups seen on the EDE. Both groups improved over time on the EDE.	⨁⨁⨁⨁ HIGH	CRITICAL
Weight (assessed with: Percent ideal body weight)
1	Case control	serious a	not serious	not serious	not serious	none	Retrospective case control study looking at MFT versus TAU for AN. 50 female adolescents aged 11–18 were included (25 in MFT group and 25 in TAU group). Those in MFT restored weight to a higher percentage (99.6% vs. 95.4%).	⨁◯◯◯ VERY LOW	CRITICAL
Weight (assessed with: kg and BMI) Psychological Symptoms (assessed with: EDE, EDI)
4	Case Series	very serious a,b	not serious	not serious	not serious	none	Four studies without a control condition. Total n = 142 adolescents (5 males, 137 females), Diagnoses were mixed including AN, EDNOS and BN. Significant improvements in weight were reported.	⨁◯◯◯ VERY LOW	CRITICAL
		very serious a,b	not serious	not serious	not serious	none	Improvements in psychological symptoms were seen pre to post MFT. In a case series of 10 adolescents aged 13 to 18 years, EDE scores decreased from 4.31 to 3.41 (cohen’s d 0.82).	⨁◯◯◯ VERY LOW	CRITICAL

^ano randomization

^bno control condition

Bibliography:

RCT – Eisler 2016 [76]

Case control - Gabel 2014 [77]

Case Series - Gelin 2015 [80], Hollesen 2013 [78], Salaminiou 2017 [79], Stewart 2015 [81]