Outcome variable Number of studies and study design Study limitations Consistency Directness Precision Publication bias Magnitude of
effect		 Absolute effect I=fixed retainer
C=removable retainer		 Quality of evidence
GRADE
Treatment stability 2 RCT 1 Cohort Some limitations (?)a No inconsistency Serious indirectness
(-1)b		 Some imprecision
(?)c		 Unlikely Not relevant I= Δ LII: 0.6 e
C=Δ LII: 1.6		 ⊕⃝⃝⃝
  Periodontal outcomes   1 RCT 3 Cohort   Serious limitations
(-1)a 		  No inconsistency   Serious indirectness
(-1)d		   Serious imprecision
(-1)c		   Unlikely   Not relevant   Not applicable f   ⊕⃝⃝⃝
  Dental Caries   1 RCT 2 Cohort   Serious limitations
(-1)a 		  No inconsistency   Serious indirectness
(-1)d		   Serious imprecision
(-1)c		   Unlikely   Not relevant   I=0
C=0		   ⊕⃝⃝⃝
  Dental plaque     1 RCT 3 Cohort   Serious limitations
(-1)a 		  No inconsistency   Serious indirectness
(-1)b		   Serious imprecision
(-1)c		   Unlikely   Not relevant   Not applicable f   ⊕⃝⃝⃝   
  Calculus   1 RCT 2 Cohort   Serious limitations
(-1)a 		  No inconsistency   Serious indirectness
(-1)d 		  Serious imprecision
(-1)c 		  Unlikely   Not relevant   Not applicable f   ⊕⃝⃝⃝
  Gingivitis   1 RCT 3 Cohort   Serious limitations
(-1)a 		  No inconsistency   Serious indirectness
(-1)d 		  Serious imprecision
(-1)c 		  Unlikely   Not relevant   Not applicable f   ⊕⃝⃝⃝
⊕⊕⊕⊕ = high quality of evidence, ⊕⊕⊕⃝ = moderate quality of evidence, ⊕⊕⃝⃝ =Low quality of evidence, ⊕⃝⃝⃝ =Low quality of evidence.[10]
a Limitations in randomization procedure, and no blinding.
b Study population strictly defined in one RCT, and undefined patient characteristics in the other RCT.
c Small groups in some studies, and no 95%CI presented.
d Patient characteristics not clearly described.
e Data from Edman Tynelius et al., 2013. LII=Little’s irregularity index.
f No pooled effect estimates, due to different outcomes across the studies.
Table 16: Summary of Findings - Fixed retainer versus removable retainer.