Physical therapy for Bell s palsy (idiopathic facial paralysis) (Review)

Transcription

1 Physical therapy for Bell s palsy (idiopathic facial paralysis) (Review) Teixeira LJ, Soares BGDO, Vieira VP, Prado GF This is a reprint of a Cochrane review, prepared and maintained by The Cochrane Collaboration and published in The Cochrane Library 2008, Issue 3

2 T A B L E O F C O N T E N T S HEADER ABSTRACT PLAIN LANGUAGE SUMMARY BACKGROUND OBJECTIVES METHODS RESULTS DISCUSSION AUTHORS CONCLUSIONS ACKNOWLEDGEMENTS REFERENCES CHARACTERISTICS OF STUDIES DATA AND ANALYSES Analysis 1.1. Comparison 1 ELECTROSTIMULATION VERSUS CONTROL, Outcome 1 Incomplete recovery after 6 and 12 months Analysis 1.2. Comparison 1 ELECTROSTIMULATION VERSUS CONTROL, Outcome 2 Mean Facial Grading Scale after 3 months Analysis 2.1. Comparison 2 ELECTROSTIMULATION VERSUS PREDNISONE, Outcome 1 Incomplete recovery after six months (all participants) Analysis 2.2. Comparison 2 ELECTROSTIMULATION VERSUS PREDNISONE, Outcome 2 Incomplete recovery six months according severity Analysis 2.3. Comparison 2 ELECTROSTIMULATION VERSUS PREDNISONE, Outcome 3 Mean time to complete recovery (in days) Analysis 3.1. Comparison 3 EXERCISE VERSUS WAITING LIST, Outcome 1 Recovery on Facial Grading Scale (Sunnybrook scale) Analysis 3.2. Comparison 3 EXERCISE VERSUS WAITING LIST, Outcome 2 Recovery on Facial Disability Indexphysical Analysis 4.1. Comparison 4 EXERCISE VERSUS CONVENTIONAL TREATMENT, Outcome 1 Incomplete recovery three months after randomisation Analysis 4.2. Comparison 4 EXERCISE VERSUS CONVENTIONAL TREATMENT, Outcome 2 Mean time from the beginning of the recovery (in weeks) Analysis 4.3. Comparison 4 EXERCISE VERSUS CONVENTIONAL TREATMENT, Outcome 3 Mean time from completion of recovery (in weeks) Analysis 4.4. Comparison 4 EXERCISE VERSUS CONVENTIONAL TREATMENT, Outcome 4 Potmann Score. 30 APPENDICES HISTORY CONTRIBUTIONS OF AUTHORS DECLARATIONS OF INTEREST INDEX TERMS i

3 [Intervention Review] Physical therapy for Bell s palsy (idiopathic facial paralysis) Lázaro Juliano Teixeira 1, Bernardo Garcia de Oliveira Soares 2, Vanessa Pedrosa Vieira 3, Gilmar F Prado 4 1 Department of Physiotherapy, Prefeitura Municipal de Balneario Camboriu, Santa Catarina, Brazil. 2 Universidade Federal de São Paulo, São Paulo, Brazil. 3 Medicina Interna e terapêutica, UNIFESP - Universidade Federal de São Paulo, São Paulo, Brazil. 4 São Paulo, Brazil Contact address: Lázaro Juliano Teixeira, Department of Physiotherapy, Prefeitura Municipal de Balneario Camboriu, R. Ana Garcia Pereira, n 167, Balneario Camboriu, Santa Catarina, , Brazil. [email protected]. Editorial group: Cochrane Neuromuscular Disease Group. Publication status and date: New, published in Issue 3, Review content assessed as up-to-date: 3 February Citation: Teixeira LJ, Soares BGDO, Vieira VP, Prado GF. Physical therapy for Bell s palsy (idiopathic facial paralysis). Cochrane Database of Systematic Reviews 2008, Issue 3. Art. No.: CD DOI: / CD pub2. Background A B S T R A C T Bell s palsy (idiopathic facial paralysis) is commonly treated by physical therapy services with various therapeutic strategies and devices. There are many questions about their efficacy and effectiveness. Objectives To evaluate the efficacy of physical therapies on the outcome of Bell s palsy. Search strategy We searched the Cochrane Neuromuscular Disease Group Trials Register (February 2008), the Cochrane Central Register of Controlled Trials (The Cochrane Library, Issue 4, 2007), MEDLINE (January 1966 to February 2008), EMBASE (January 1980 to February 2008), LILACS (January 1982 to February 2008), PEDro (from 1929 to February 2008), and CINAHL (January 1982 to February 2008). Selection criteria We selected randomised or quasi-randomised controlled trials involving any physical therapy. We included participants of any age with a diagnosis of Bell s palsy and all degrees of severity. The outcome measures were: incomplete recovery six months after randomisation, motor synkinesis, crocodile tears or facial spasm six months after onset, incomplete recovery after one year and adverse effects attributable to the intervention. Data collection and analysis Titles and abstracts identified from the register were scrutinized. The assessment of methodological quality took into account secure method of randomisation, allocation concealment, observer blinding, patient blinding, differences at baseline of the experimental groups, and completeness of follow-up. Data were extracted using a specially constructed data extraction form. Separate subgroup analyses of participants with more and less severe disability were undertaken. 1

4 Main results The search identified 45 potentially relevant articles. Six studies met the inclusion criteria. Three trials studied the efficacy of electrostimulation (294 participants) and three exercises (253 participants). Neither treatment produced significantly more improvement than the control treatment or no treatment. There was limited evidence that improvement began earlier in the exercise group. Authors conclusions There is no evidence of significant benefit or harm from any physical therapy for idiopathic facial paralysis. The possibility that facial exercise reduces time to recover and sequelae needs confirming with good quality randomised controlled trials. P L A I N L A N G U A G E S U M M A R Y Physical treatments for idiopathic facial paralysis Bell s palsy is an acute disorder of the facial nerve, which produces full or partial loss of movement on one side of the face. The facial palsy gets completely better without treatment in most, but not all, people. Physical therapies, such as exercise, biofeedback, laser, electrotherapy, massage and thermotherapy, are used to hasten recovery. This review of existing trials found insufficient evidence to decide whether any of these therapies work. More trials are needed to assess their effects. B A C K G R O U N D Idiopathic facial palsy, also called Bell s palsy, is an acute disorder of the facial nerve, which may begin with symptoms of pain in the mastoid region and produce full or partial paralysis of movement of one side of the face (Adour 1982; Valença 2001). Its cause is not known (Peitersen 2002). Increasing evidence suggests that the main cause of Bell s palsy is reactivation of latent herpes simplex virus type 1 in the cranial nerve ganglia (De Diego 1999; Holland 2004; Valença 2001). How the virus damages the facial nerve is uncertain (Gilden 2004). The annual incidence of Bell s palsy varies widely, ranging between 11.5 and 40.2 cases per 100,000 population (De Diego 1999; Peitersen 2002). There are peaks of incidence in the 30 to 50 and 60 to 70 year old age groups (Gilden 2004; Gonçalvez 1997). Bell s palsy has a fair prognosis without treatment (Holland 2004). According to Peitersen (Peitersen 2002), complete recovery was observed in 71% of all patients. Ninety-four per cent of patients with incomplete and 61% with complete paralysis made a complete recovery. The main question is whether results would be better if some treatment were given. About 23% of people with Bell s palsy are left with either moderate to severe symptoms, hemifacial spasm, partial motor recovery, crocodile tears (tears upon salivation), contracture or synkinesis (involuntary twitching of the face or blinking). Recurrence occurs in about 8.3% (Valença 2001). The prognosis depends to a great extent on the time at which recovery begins. Early recovery gives a good prognosis and late recovery a bad prognosis. If recovery begins within one week, 88% obtain full recovery, within one to two weeks 83% and within two to three weeks 61%. Normal taste, stapedius reflex and tearing give a significantly better prognosis than if these functions are impaired. Recovery is less likely to be satisfactory with complete rather than incomplete paralysis, with pain behind the ear and in older people (Danielidis 1999). Other poor prognostic factors include hypertension and diabetes mellitus (Gilden 2004; Peitersen 2002). Evaluation of therapy is made difficult because of the high rates of spontaneous and complete recovery (Peitersen 2002). The principles of treatment in the acute phase have not changed over the past 20 years (Adour 1982). They focus on protection of the cornea from drying and abrasion due to impaired lid closure and tear production. Lubricating drops are recommended during the day and a simple eye ointment at night (Holland 2004; Valença 2001). 2

5 Cochrane reviews concluded that the available evidence did not show significant benefit from acyclovir or similar agents (Allen 2004), steroid therapy (Salinas 2004) or acupuncture (He 2007). However the Sullivan 2007 study with 496 participants compared different combinations of prednisolone, acyclovir and placebo. They found significant benefit from prednisolone but not acyclovir. Hato 2007 assessed the efficacy of valacyclovir with 296 participants divided into two groups (valacyclovir with prednisolone, and placebo with prednisolone) and found significant benefit from valacyclovir. Some authors suggest that facial nerve decompression be considered, although there are no data from clinical trials to support its use (Adour 2002; Gilden 2004; Grogan 2001; Peitersen 2002). Thermal methods, electrotherapy (which uses an electrical current to cause a single muscle or group of muscles to contract), massage, facial exercises and biofeedback are forms of physical therapy that have been used (Mosforth 1958; Peitersen 2002). Exercise therapy has been used more than other interventions (Beurskens 2003; Brach 1999; Ross 1991; Segal 1995a). In the last few years other systematic reviews have been undertaken. Beurskens 2004 searched electronic databases and included two studies (Ross 1991; Segal 1995a) which did not show a significant effect of intervention. Quinn 2003 searched for electrotherapy interventions in electronic databases, reference lists in the studies and contacted experts from 1975 to They concluded that the benefit of electrotherapy was unclear due to inadequate research methods, sample sizes and dose information. Despite this, they provide a very good discussion about the current knowledge of the anatomy, physiology and pathomechanics during the course of the palsy to support the use of physiotherapy resources. None of the trials considered in these reviews fulfilled the criteria for this review. Peitersen 2002 also highlighted the lack of evidence for current treatments, for thermal methods (conductive, radiative and convective heat transfer in order to achieve vasodilatation or ice over the mastoid region with the aim of relieving oedema), electrotherapy, massage and facial exercise. O B J E C T I V E S The objective of this systematic review was to evaluate the efficacy of physical therapies for Bell s palsy (idiopathic facial palsy). M E T H O D S Criteria for considering studies for this review Types of studies We included all randomised or quasi-randomised (alternate or other systematic allocation) controlled trials involving any physical therapy compared with no treatment, placebo treatment, drug treatment, acupuncture or other physical therapy interventions. Types of participants We included participants with a diagnosis of Bell s palsy, defined as idiopathic lower motor neuron facial palsy of sudden onset. Participants of any age, and all degrees of severity were included. People with facial palsy due to Ramsay-Hunt syndrome or other recognised causes were not included. Types of interventions We included trials of any form of physical therapy treatment compared with either no treatment or drugs or an alternative form of non-drug treatment. Physical therapy was considered as the use in treatment of any physical agents, such as heat, light, cold, sound, water, electricity, manual therapy and other gadgets working on physical principles. Types of physical therapy interventions for facial palsy included facial exercises, such as strengthening and stretching, endurance, therapeutic and facial mimic exercises ( mime therapy ) (Beurskens 2003), electrotherapy, biofeedback, transcutaneous electrical nerve stimulation (TENS) or electrical neural muscular stimulation (ENMS), thermal methods or massage, alone or in combination with any other therapy. Types of outcome measures The primary outcome measure was incomplete recovery six months after randomisation. Incomplete recovery was defined in two ways. Participants who had House Grade III (moderate dysfunction) or worse (House 1985) at entry were considered to have incomplete recovery if they still had House Grade III or worse. For participants who had House Grade II at entry, incomplete recovery was defined as a persistent House Grade II or worse after six months. If the House Grade score was not available, another similar facial nerve disability score was used instead (House 1985; VanSwearingen 1996). Secondary outcome measures were: 1. the presence of motor synkinesis, contracture, hyperkinesia, facial spasm or crocodile tears six months after onset; 2. complete or incomplete recovery after one year; 3. adverse effects attributable to the intervention such as pain or worsening of condition. Search methods for identification of studies We searched the Cochrane Neuromuscular Disease Group Trials Register in February 2008 using the terms Bell s palsy or 3

6 idiopathic facial paralysis or facial palsy. We also searched the Cochrane Central Register of Controlled Trials (The Cochrane Library, Issue 4, 2007), MEDLINE (January 1966 to February 2008), EMBASE (January 1980 to February 2008), LILACS (January 1982 to February 2008), and CINAHL (January 1982 to February 2008) and PEDro (from 1929 to February 2008). Electronic searches See Appendix 1, Appendix 2, Appendix 3, Appendix 4, Appendix 5. Searching other resources 1. We checked references of all identified trials. 2. We contacted physical therapy companies in order to obtain data on unpublished trials. 3. We contacted first authors of all included trials for further information or information regarding unpublished trials. Data collection and analysis Study selection Two authors (LJT, VPV) scrutinized titles and abstracts identified from the register. The full texts of all potentially relevant studies were obtained for independent assessment by the authors. Two authors decided which trials fitted the inclusion criteria. Disagreements about inclusion criteria were resolved by consensus and consultation with a third author (BGOS). Assessment of methodological quality The assessment of methodological quality took into account secure method of randomisation, allocation concealment, observer blinding, patient blinding, differences at baseline of the experimental groups, and completeness of follow-up. These items were assessed according to the Cochrane Collaboration standard scheme: grade A: adequate, grade B: unclear, grade C: inadequate or not done. Two authors (LJT, VPV) assessed quality independently. Disagreement between the authors was resolved by discussion if necessary with a third author (BGOS). Data extraction Two authors independently extracted data on participants, methods, interventions, outcomes and results using a specially constructed data extraction form. Missing data were obtained from the trial authors whenever possible. Analysis of data Data were entered and analysed using Review Manager (RevMan) software. For dichotomous data, relative risks (RR) with 95% confidence intervals (CI) were estimated based on the fixedeffect model or on the random effects model if heterogeneity was present. The number needed to treat (NNT) and number needed to harm (NNH) were calculated if possible. For continuous outcomes, weighted mean differences (WMD) between groups were estimated. Heterogeneity was assessed by the chi-squared test and was assumed to be present when the significance level was lower than 0.10 (p < 0.10). When significant heterogeneity was present, an attempt was made to explain the differences based on clinical characteristics of the included studies. A sensitivity analysis was performed, omitting trials which included participants with different clinical characteristics or trials with lower methodological quality. If there had been sufficient trials of the same intervention, we would have constructed a funnel plot (of trial effect versus trial size) to assess potential publication bias. Subgroup analysis Separate subgroup analyses of participants with more severe disability (House Grade III or worse) and less severe disability (House Grade II or better) were undertaken. We also considered patients treated before and after two weeks from onset. R E S U L T S Description of studies See: Characteristics of included studies; Characteristics of excluded studies; Characteristics of studies awaiting classification. The literature search and hand searching identified 41 potentially relevant articles (Cochrane databases = 7, MEDLINE = 17, EM- BASE = 23, CINAHL = 6, LILACS = 3; PEDro = 11 and handsearching = 2), (see Figure 1). Some studies were found in more than one database. Of these, six trials met the inclusion criteria ( Beurskens 2003; Flores 1998; Manikandan 2007; Mosforth 1958; Wang 2004; Wen 2004). Eight other studies, all in Chinese, await translation and assessment. Excluded studies Twenty seven studies were excluded because they were: 1. Series of cases or case reports (Aleev 1973; Brach 1999; Brown 1978; Coulson 2006b; Danile 1982; Lobzin 1989; Manca 1997; Romero 1982; Segal 1995a). 2. Retrospective studies (Bernardes 2004; Cronin 2003; Dalla- Toffola 2005). 4

7 3. Non-systematic reviews (Goulart 2002; Beurskens 2004). 4. Not physical therapy interventions (Casler 1990; Klingler 1982; Taverner 1966; Zhao 2005). 5. Constrained by methodological restrictions (Dubravica 1996; Koyama 2005; Murakami 1993; Nakamura 2003; Shiau 1995). 6. Composed of few participants with idiopathic facial palsy ( Balliet 1982; Coulson 2006; Ross 1991; Segal 1995b).Details can be found in the Characteristics of excluded studies. Included studies The six studies (Beurskens 2003; Flores 1998; Manikandan 2007; Mosforth 1958; Wang 2004; Wen 2004) (See Characteristics of included studies) included a total of 547 people. Three trials studied manual therapy and electrostimulation (Flores 1998; Manikandan 2007; Mosforth 1958) (294 participants), and three involved exercise (Beurskens 2003; Wen 2004; Wang 2004) (253 participants). The first study evaluating physical therapy for facial palsy was one of the first physical therapy randomised controlled trials described for any condition. Mosforth (Mosforth 1958) studied 86 people with acute Bell s palsy of less than 14 days duration. Three participants were lost to follow up. Auto-massage of the face, infrared and interrupted galvanic stimulation (pulse 100 msec) in 44 participants, was compared to massage alone in 42 participants. Treatment was continued until recovery or until the condition seemed stationary (two to six months). The outcomes were electrical examination and grade of paralysis estimated visually as a percentage of the function of the normal side, the time to begin improvement and the time to complete recovery. Manikandan 2007 assessed the results of 59 participants with acute facial palsy and compared two groups with physiotherapy interventions. Although the objective of the study was to test a specific exercise strategy and both groups undertook different exercises, the regimen adopted was similar (home based exercises) and electrotherapy was the main difference between the groups. One group of 30 people underwent a fixed protocol with electrostimulation (galvanic and faradic currents) for the two first weeks, massage and gross facial exercises. The other 29 people learnt an individualized exercise program focused on the quality of the exercises and not on the quantity. The movements were to be symmetrical without voluntary movement of the uninvolved side. All individuals were assessed by the Facial Grading Scale (Roos 1996) at the outset and after three months. In Flores 1998, there were 149 participants with acute Bell s palsy (onset in one to three days). Twenty-nine people (19.46%) dropped out without a description of the reason for drop out. One group of 76 people were treated with infrared treatment and electrostimulation and were compared to 72 people treated with prednisone for up to 14 days. Outcomes were time to recover, clinical history and a functional scale (May Scale). Authors analysed different groups according to whether the lesion was thought to be proximal or distal to the origin of the chorda tympani nerve. The Flores 1998 study was analysed with caution. Because corticosteroids have now been shown to be efficacious (Sullivan 2007), comparing physical therapies with this active treatment could be considered inappropriate. Nevertheless we included this study and discussed some outcomes. Beurskens 2003 studied 50 people with chronic (more than nine months) facial paralysis. Only 34 people had idiopathic facial palsy. Sixteen received exercises (mime therapy) and the other 18 formed a waiting list control group. Mime therapy consists mainly of facial mimic exercises. Outcomes were face stiffness, lip mobility, the Facial Disability Index (VanSwearingen 1996), the Sunnybrook Facial Grading Scale (Roos 1996), and the House-Brackmann Facial Grading System (House 1985). The author kindly sent us all the outcomes for the idiopathic facial palsy participants. The mean baseline House-Brackmann score was 4 and after one year it was 3 for all the treatment participants. Although all participants apparently improved, in the protocol for this review we made the assumption that a grade over 3 could mean improvement but does not mean recovery. It was based on a previous study (Peitersen 2002) and the clinical meaning of House-Brackman grades (House 1985). The continuous data results were more significant. However, the samples were composed of 16 and 18 individuals with Bell s palsy in the exercise and control groups respectively. This small sample is a significant limitation. More observations on this study are made in the Discussion. Wen 2004 studied 145 people with acute idiopathic facial palsy for 12 weeks. Eighty-five participants were submitted to a combination of conventional therapy plus facial rehabilitation exercises (movements using facial muscles) while 60 participants received only conventional therapy not detailed in the translation process. This Chinese study presented the following outcomes analysed in the review: (1) time when the patient started to recover and (2) time that the recovery was complete. The study analysed groups of mild, moderate and severe dysfunction patients. Wang 2004 treated 74 people with acute Bell s palsy with two different strategies. Both groups received medicine (cortisone, and mexobalamin and vitamin B2), physical treatment (not described in the translation), massage, and acupuncture. For the exercise therapy (n = 43), functional exercises were added. The outcome was facial muscle function with the Potmann Score after one month. Risk of bias in included studies The scores of methodological quality for each trial can be found in Table 1. 5

8 Table 1. Methodological Quality Assessment Study ID Secure method random Allocation Blinding Follow-up Mosforth 1958 A A Not done A Flores 1998 B B Not done B Wen 2004 B B Not done C Wang 2004 A A Assessor blind A Beurskens 2006 C C Assessor blind A Manikandan 2007 A A Not done A Secure method of randomisation and allocation concealment Mosforth 1958 randomised the groups using a prepared list (grade A). Beurskens (Beurskens 2003) used a coin flip to assign the first participant and the others were allocated by alternation and this was classified as inadequate allocation concealment (grade C). Manikandan 2007 used a method of six blocks with 10 in each block (grade A) and Wang 2004 randomised their samples by computer (grade A). Two studies (Flores 1998; Wen 2004) classified their trial as randomised, but they did not describe the method of randomisation (grade B). Blinding Due to the nature of the intervention evaluated in this review, effective blinding of the participants is problematic. Placebo electrostimulation could have been used but blinding to exercise interventions is impractical or impossible. Blinding of outcome assessors can be achieved but only Beurskens 2003 and Wang 2004 blinded the assessor. Completeness of follow-up Beurskens 2003 and Mosforth 1958 assessed outcomes at one year follow up. Beurskens analysed all the 50 participants (of whom 34 had Bell s palsy) at this time. Mosforth did not analyse all the people after one year because they were discharged when recovered. Despite this, our analyses were not affected. The data of interest (incomplete recovery) were reported by the study and drop outs were considered to have incomplete recovery in our intention to treat analysis. Flores 1998 did not describe follow up and 29 people (19.26%) dropped out without description of their allocation groups. The reasons described were that the participants requested another medication or they had not adhered to the treatment. Manikandan 2007 and Wen 2004 followed subjects until 12 weeks and Wang 2004 until 30 days, the end of the treatment period. Effects of interventions The results have been divided by the intervention and described according to the results for each of our outcome measures. ELECTRICAL STIMULATION Differences in baseline between groups Wen 2004 did not present the baseline characteristics of the participants. Manikandan 2007, Mosforth 1958, Flores 1998, and Wang 2004 reported number, sex, age and duration of the palsy indicating no significant differences between groups. Beurskens 2003 reported no significant differences between groups at baseline for demographic data and severity and duration of facial palsy. Primary outcome measure Mosforth 1958 studied the efficacy of electrotherapy after six months in a total of 86 participants (n = 44 electrical stimulation and n = 42 control). The graphs were constructed using an intention-to-treat analysis and less than 75% recovery was considered 6

9 a bad outcome. The relative rate of improvement was not significantly different, relative risk (RR) 1.30, 95% CI 0.68 to 2.5 (see Analysis 1.1). Manikandan 2007 described results after three months on a continuous scale. The Facial Grade Score measured rest score, synkinesis scores and movement score of the 28 participants in each group. The first two scores did not show statistical significance. The movement score improved significantly in the group without electrical stimulation, mean difference (MD) 68.00, 95% CI to (see Analysis 1.2). Consequently the total score improved, MD 12.00, 95% CI 1.26 to (see Analysis 1.2). In Flores 1998 study ten (12.98%) of the 77 participants that were treated with electrical stimulation, and 11 (15.27%) of the 72 treated with prednisone had incomplete recovery after 6 months, RR 0.85, 95% CI 0.38 to 1.88 (see Analysis 2.1). Secondary outcome measures (1) Presence of motor synkinesis, contracture, hyperkinesia, facial spasm or crocodile tears six months after onset Mosforth 1958 showed no significant differences between the group receiving electrical stimulation and the control group in respect of facial muscle contracture. Eleven participants (25%) in the treated group and eight (20%) in the control had contracture, RR 1.25, 95% CI 0.56 to Manikandan 2007 reported 2 participants in the group receiving exercise and electrical stimulation that presented with mild synkinesis after three months, and none in the group with exercise alone. This was considered non significant, RR 0.20, 95% CI 0.01 to (2) Incomplete recovery after one year Mosforth 1958 reported no statistically significant differences in this assessment, RR 1.15, 95% CI 0.55 to 2.36 (see Analysis 1.1). (3) Adverse effects attributable to the intervention such as pain or worsening of condition No adverse effects were attributed to the interventions. Subgroup analyses Flores 1998 undertook a subgroup analysis by severity of the axonal damage. In the group with mild disease or with lesions distal to the chorda tympani lesion (n = 102) all individuals in both groups improved at six months. In the most severe group or lesions proximal to the chorda tympani (n = 47) there was no significant difference in recovery, RR = 0.62, 95% CI 0.34 to 1.15 (see Analysis 2.2). Analysing mean time to recovery of the 149 participants in the study in days, we found significantly faster recovery with electrical stimulation, (MD -8.38, 95% CI to Analysis 2.3). However Manikandan 2007 gave opposite results. FACIAL EXERCISES Primary outcome measure Beurskens 2003 included only 34 participants with chronic (more than nine months) Bell s palsy. All participants in the exercise group improved but none in the control group improved. Wen 2004 compared facial exercises (n = 85) with medication (n = 60). There was no significant difference in improvement between the groups at three months % of participants in the exercise group and 88.33% of participants in the control group recovered, RR % CI 0.21 to 1.71 (see Analysis 4.1). Wang 2004 compared a combination of medicines, acupuncture and physiotherapy (n = 31) with the same interventions plus functional exercises (n = 43). The single outcome was facial muscle function (Potmann Score) after one month. It showed a statistically significant difference in favour of the functional exercise group, MD 8.47, 95% CI 7.05 to Secondary outcome measures (1) Presence of motor synkinesis, contracture, hyperkinesia, facial spasm or crocodile tears six months after onset Wen 2004 reported significantly less facial motor synkinesis after exercise, with 12 cases in the control group (20%) and 4 cases in the exercise group (4.7%), RR = 0.24, 95% CI 0.08 to (2) Incomplete recovery after one year According to the criteria for this review, none of the participants recovered completely after one year. Beurskens 2003 kindly sent us the continuous outcomes after one year. These all showed improvements in favour of the exercise group: Facial Grading Scale WMD 20.40, 95% CI 8.74 to 32.04; Facial Disability Index Physical MD 10.30, 95% CI to 21.97, and the Facial Disability Index Social MD 14.50, 95% CI 4.85 to (see Analysis 3.1, Analysis 3.2 ), in favour of exercise. (3) Adverse effects attributable to the intervention such as pain or worsening of condition No adverse effects were attributable to the interventions 7

10 Subgroup analyses Wen 2004 presented data on participants with mild and more severe disease. There was no difference in the proportion of participants that improved in the exercise group and conventional therapy group in the individuals with mild paralysis (Analysis 4.2). But when we analysed the sub-group with moderate severity, we observed that the exercise group began (Analysis 4.2) and finished (Analysis 4.3) improving sooner. D I S C U S S I O N In the light of the numerous physical therapies used for treating Bell s palsy in daily practice, this review highlights the lack of high quality evidence to support the use of these strategies. Electrotherapy, exercises, biofeedback, manual therapy and laser were evaluated in some studies, but only trials involving electrostimulation and exercise had the minimum methodological quality to be considered for this systematic review. Electrical stimulation Almost all the outcomes reported failed to show any statistically significant difference between either electrotherapy or exercises and conventional or no treatment. Mosforth 1958 concluded that it is not possible to recommend electrostimulation and questions its cost-effectiveness. The results of Manikandan 2007 are in agreement as the group with electrical stimulation had worse quality of movement and functional recovery after three months. Flores 1998 found no differences in the proportion of participants with recovery after six months. The time to recovery in the Flores study was less in the electrostimulation group but the study had some methodological restrictions such as comparing physiotherapy with prednisone, an active treatment, and almost 20% participant drop outs. No statistical differences were found in synkineses or other complications in any of the trials. Exercise Neither Wen 2004 studying acute cases nor Beurskens 2003 studying chronic cases found differences in the proportion recovering after three and six months. Significantly less synkinesis was observed by Wen 2004 after three months. The evidence was limited by the restrictions of reported outcomes to continuous data. The assessment was blinded in two studies (Beurskens 2003; Wang 2004). Comments about the methodology Almost all the included studies had some limitations to be considered in future research. In the electrical stimulation trials Flores 1998 compared electrostimulation and prednisolone, an active treatment, which could have biased the study results. Manikandan 2007 used different exercise regimens in both groups but the main difference was the use of electrical stimulation in one of the groups. This modified the way data have been analysed and we considered that the study tested electrical stimulation rather than different exercise regimens. In the exercise trials, Beurskens 2003 studied chronic facial palsy and included participants with dysfunctions other than idiopathic facial palsy, which reduced the size of the sample of interest for this review and limited conclusions. Wang 2004 and Wen 2004 compared combinations of physiotherapy and medicine with functional exercises which complicated interpretation. The main outcomes used were continuous scales of motor function. In another publication Beurskens 2004b discussed the outcomes of applying exercise to treat facial paresis. He observed a significant recovery in the outcomes in participants receiving exercise for palsies lasting more than nine months: asymmetry in the face at rest, asymmetry during voluntary facial movements, synkineses, complaints concerning pain, stiffness, involuntary movements, reports concerning difficulties in eating, drinking, speaking, and patient perception about their quality of life. Although the House-Brackmann score was used as an overall measure of facial impairment, the authors stated that it was not sensitive enough to measure improvement during therapy with exercise in chronic cases. The Facial Grading Scale (Roos 1996) and the Facial Disability Index (VanSwearingen 1996) were considered good assessment options. We would have preferred to convert continuous data into dichotomous data. For example, for recent Bell s palsy we expect a minimum of 71% recovery (House-Brackmann scores of I or II) before three months. In chronic stationary cases with House-Brackman scores of III or IV, patients might find lesser degrees of improvement valuable. In subgroups with severe dysfunction, complications or sequelae were the clinical outcomes considered. Peitersen 2002 reported that out of more than 2500 people with facial paralysis, 29% had persistent weakness, 17% contracture and 16% synkinesis. Wen 2004 described twelve cases out of 85 participants with synkinesis in the control group (14%) and four out of 60 cases in the exercise group (6.6%). More studies are needed to confirm this. However, the trials in which improvements were reported as continuous outcomes are less reliable, particularly if they were not blinded. It is not impossible to blind such studies, since the authors can either introduce an observer who had not seen the patient before or take photographs or even videos, as in the Beurskens 2003 and Wang 2004 studies. Nevertheless these studies had other limitations. 8

11 Other clinical references used in the studies were the times to onset of recovery and times to complete recovery. Some differences emerged between the groups treated with physical therapy and other treatment (prednisone or other medication). The time to improvement seemed to be shorter in participants receiving physical therapies, even in mild, as well as moderate grades of paralysis but these results are really not reliable. The conclusions of this systematic review were limited by the low number and poor quality of studies and the heterogeneity of the results. including dose and duration. Outcome measures should be selected which are likely to be adequately responsive for detecting change with physical therapies. Measures should include facial appearance, function (eating and drinking and speaking), facial appearance (including asymmetry and involuntary movements), and quality of life. Recovery at defined times such as three, six and twelve months of treatment is easier to measure accurately than the time to recovery. Use of photography or video to blind the outcome assessor is encouraged. A U T H O R S C O N C L U S I O N S Implications for practice There is no evidence of significant benefit or harm from the limited trials of electrical stimulation or facial exercises for Bell s palsy. Implications for research There is a need for well-designed, randomised trials of electrical stimulation, exercise and other physical therapies for Bell s palsy. Reports of such trials should give details of the treatments given A C K N O W L E D G E M E N T S The staff of the Brazilian Cochrane Centre, Cochrane Neuromuscular Desease Group, Dr David Allen for his important contribution reviewing the protocol and a special thanks to Professor Richard Hughes for all the comments during all the editorial process. Rachel Barton for the search strategy and the database searches. To Dr Zhannat Idrissova, Dr Hitoshi Nukada and Yuquian Ma for the translations. Kate Jewitt, Janice Fernandes and Jane Batchelor for all the support. Special thanks to my wife, Cinira Gomes, and our daughter Rafaela for the patience and love all the time. R E F E R E N C E S References to studies included in this review Beurskens 2003 {published and unpublished data} Beurskens CHG, Heymans PG. Mime therapy improves facial symmetry in people with long-term facial nerve paresis: a randomized controlled trial. Australian Journal of Physiotherapy 2006;52(3): Beurskens CHG, Heymans PG. Positive effects of mime therapy on sequelae of facial paralysis: stiffness, lip mobility, and social and physical aspects of facial disability. Otology & Neurotology 2003;24 (4): Beurskens CHG, Heymans PG, Oostendorp RAB. Stability of benefits of mime therapy in sequelae of facial nerve paresis during a 1-year period. Otology & Neurotology 2006;27(7): Flores 1998 {published data only} Flores PF, Medina RZ, Haro LG. Idiopathic peripheral facial paralysis treatment physic therapy versus prednisone [Tratamiento de la parálisis facial periférica idiopática: terapia física versus prednisona]. Revista médica del Instituto Mexicano del Seguro Social 1998;36(3): Manikandan 2007 {published and unpublished data} Manikandan N. Effect of facial neuromuscular re-education on facial symmetry in patients with Bell s palsy: a randomized controlled trial. Clinical Rehabilitation 2007;21(4): Mosforth 1958 {published data only} Mosforth J, Taverner D. Physiotherapy for Bell s palsy. British Medical Journal 1958;2(5097): Wang 2004 {unpublished data only} Wang XH, Zhang LM, Han M, Zhang KQ. Clinical application of functional exercise and staged therapy in treatment of facial nerve paralysis. Zhonghua Linchuang Kangfu Zazhi [Chinese Journal of Experimental and Clinical Virology] 2004;8(4): Wen 2004 {unpublished data only} Wen CM, Zhang BC. Effect of rehabilitation training at different degree in the treatment of idiopathic facial palsy: a randomized controlled comparison. Zhongguo Linchuang Kangfu 2004;8(13): References to studies excluded from this review Aleev 1973 {unpublished data only} Aleev LS. Experience in the treatment of facial nerve neuritis using the method of programmed multi-channel bioelectrical control. Zhurnal Nevropatologii i Psikhiatrii Imeni S. S. Korsakova 1973;73 (3): Balliet 1982 {published data only} Balliet R, Shinn JB, Bach-y-Rita P. Facial paralysis rehabilitation: retraining selective muscle control. International Rehabilitation Medicine 1982;4(2):

12 Bernardes 2004 {published data only} Bernardes DFF, Gomez MVSG, Pirana S, Bento RF. Functional profile in patients with facial paralysis treated in a myofunctional approach. Pró-fono 2004;16(2): Beurskens 2004c {published data only} Beurskens CHG, Devriese PP, van Heiningen I, Oostendorp RAB. The use of mime therapy as a rehabilitation method for patients with facial nerve paresis. International Journal of Therapy and Rehabilitation 2004;11(5): Brach 1999 {published data only} Brach JS, VanSwearingen JM. Physical therapy for facial paralysis: a tailored treatment approach. Physical Therapy 1999;79(4): Brown 1978 {published data only} Brown DM, Nahai F, Wolf S, Basmajian JV. Electromyographic biofeedback in the reeducation of facial palsy. American Journal of Physical Medicine 1978;57(4): Casler 1990 {published data only} Casler JD, Conley J. Simultaneous dual system rehabilitation in the treatment of facial paralysis. Archives of Otolaryngology Head and Neck Surgery 1990;116(12): Coulson 2006 {published data only} Coulson SE, Adams RD, O Dwyer NJ, Croxson GR. Physiotherapy rehabilitation of the smile after long-term facial nerve palsy using video self-modeling and implementation intentions. Otolaryngology - Head and Neck Surgery 2006;134(1): Coulson 2006b {published data only} Coulson SE, Adams RD, O Dwyer NJ, Croxson GR. Use of video self-modelling and implementation intentions following facial nerve paralysis. International Journal of Therapy and Rehabilitation 2006;13(1):30 5. Cronin 2003 {published data only} Cronin GW, Steenerson RL. The effectiveness of neuromuscular facial retraining combined with electromyography in facial paralysis rehabilitation. Otolaryngology - Head and Neck Surgery 2003;128 (4): Dalla-Toffola 2005 {published and unpublished data} Dalla-Toffola E, Bossi D, Buonocore M, Montomoli C, Petrucci L, Alfonsi E. Usefulness of BFB/EMG in facial palsy rehabilitation. Disability and Rehabilitation 2005;27(14): Danile 1982 {published data only} Danile V, Marongiu A, Candioto G. New therapeutic method by ionophoresis of a drug cocktail in facial paralysis a frigore [Nuovo metodo terapeutico ionoforetico con cocktail medicamentoso della paresi facciale a frigore]. Policlinico - Sezione Medica 1982;89(2): Dubravica 1996 {published data only} Dubravica M, Musura M, Nesek-Madaric V, Stajner-Katusic S, Horga D. Treatment of facial palsy by EMG biofeedback technique - Muscle relaxation technique. Acta Clinica Croatica 1996;35(1-2): Goulart 2002 {published data only} Goulart F, Vasconcelos KSS, Souza MRV, Pontes PB. Physical therapy for facial paralysis using the biofeedback [A utilização do biofeedback no tratamento da paralisia facial periférica]. Acta Fisiátrica 2002;9(3): Klingler 1982 {published data only} Klingler D, Bibl D. Peripheral facial paralysis-role of early onset of therapy. Wiener Medizinische Wochenschrift 1982;132: Koyama 2005 {unpublished data only} Koyama S, Okada K, Yamakawa T, Kubo M, Amatsu H. The usefulness of manipulative physiotherapy in treating bell s palsy. Medical Journal of Minami Osaka Hospital 2005;53(1):55 7. Lobzin 1989 {unpublished data only} Lobzin VS, Smetankin AA, Tsatskina ND, Iashin NS. Treatment of Bell s palsy by using portable biofeedback devices. Zhurnal Nevropatologii i Psikhiatrii Imeni S. S. Korsakova 1989;89(10): Lobzin VS, Tsatskina ND. The adaptive biological control system with electromyographic feedback in the treatment of Bell s palsy [Russian]. Zhurnal Nevropatologii i Psikhiatrii Imeni S. S. Korsakova 1989;89(5):54 7. Manca 1997 {published data only} Manca M, Contenti E, Mura G, Basaglia N, Cavazzini L. EMG biofeedback in peripheral facial nerve palsy rehabilitation. Europa Medicophysica 1997;33(3): Murakami 1993 {published data only} Murakami F, Kemmotsu O, Kawano Y, Matsumura C, Kaseno S, Imai M. Diode low reactive level laser therapy and stellate ganglion block compared in the treatment of facial palsy. Laser Therapy 1993;5(3): Nakamura 2003 {published data only} Nakamura K, Toda N, Sakamaki K, Kashima K, Takeda N. Biofeedback rehabilitation for prevention of synkinesis after facial palsy. Otolaryngology - Head and Neck Surgery 2003;128(4): Romero 1982 {published data only} Corral-Romero MA, Bustamante-Balcarcel A. Biofeedback rehabilitation in seventh nerve paralysis. The Annals of Otology, Rhinology, and Laryngology 1982;92(2 Pt 1): Ross 1991 {published data only} Ross B, Nedzelski J, Mclean J. Efficacy of feedback training in longstanding facial paresis. Laryngoscope 1991;101(7 Pt 1): Segal 1995a {published data only} Segal B, Hunter T, Danys I, Freedman C, Black M. Minimizing synkinesis during rehabilitation of the paralyzed face: Preliminary assessment of a new small-movement therapy. Journal of Otolaryngology 1995;24(3): Segal 1995b {published data only} Segal B, Zompa L, Danys I, Black M, Shapiro M, Melmed C, et al.symmetry and synkinesis during rehabilitation of unilateral facial paralysis. Journal of Otolaryngology 1995;24(3): Shiau 1995 {published data only} Shiau J, Segal B, Danys I, Freedman R, Scott S. Long-term effects of neuromuscular rehabilitation of chronic facial paralysis. The Journal of Otolaryngology 1995;24(4): Taverner 1966 {published data only} Taverner D, Fearnley ME, Kemble F, Miles DW, Peiris OA. Prevention of denervation in Bell s palsy. British Medical Journal 1966;5484:

13 Zhao 2005 {published data only} Zhao Y, He L, Zhang QH. Effectiveness of three different treatments for peripheral facial paralysis. Chinese Journal of Clinical Rehabilitation 2005;9(29):41 3. References to studies awaiting assessment Diao 2002 {published data only} Diao L, et al.comparison of the efficacy between acupuncture and manipulation for Bell s palsy. Journal of Huaihua Medical College 2002;1(2):34 5. Guo 2006 {published data only} Guo QH, Yan JZ, Yan WS, Xiao MZ. Observation on non-invasive electrode pulse electric stimulation for treatment of Bell s palsy. Zhongguo Zhenjiu 2006;26(12): Li 2005 {published data only} Li J. Comparison the efficacy between acupuncture and manipulation for Bell s palsy. Chinese Clinical Medicine Research 2005;11(12): Pan 2004 {published data only} Pan L. Acupuncture plus short wave for 38 peripheral facial paralysis. Journal of Clinical Acupuncture & Moxibustion 2004;20 (4):26 7. Qu 2005 {published data only} Qu Y. Clinical observation on acupuncture by stages combined with exercise therapy for treatment of Bell palsy at acute stage. Zhongguo zhen jiu [Chinese Acupuncture & Moxibustion] 2005;25(8): Wang 2004b {published data only} Wang XH, Zhang LM, Han M, Zhang KQ, Jiang JJ. Treatment of Bell s palsy with combination of traditional Chinese medicine and western medicine. Hua xi kou qiang yi xue za zhi [West China Journal of Stomatology Stomatology] 2004;22(3): Yang 2001 {published data only} Yang G. Comparison of the efficacy between acupuncture and therapy apparatus for Bell s palsy. Journal of Clinical Acupuncture & Moxibustion 2001;17(8):28 9. Zhang 2005 {published data only} Zhang H. Acupuncture combined with facial muscle training for peripheral facial paralysis. Chinese Journal of Rehabilitation Theory and Practice 2005;11(12): Additional references Adour 1982 Adour KK. Current concepts in neurology: diagnosis and management of facial paralysis. The New England Journal of Medicine 1982;307(6): Adour 2002 Adour KK. Decompression for Bell s palsy: why I don t do it. European Archives of Otorhinolaryngology 2002;259(1):40 7. Allen 2004 Allen D, Dunn L. Aciclovir or valaciclovir for Bell s palsy (idiopathic facial paralysis). Cochrane Database of Systematic Reviews 2004, Issue 3. Beurskens 2004 Beurskens CHG, Burgers-Bots IAL, Kroon DW, OOstendorp RAB. Literature review of evidence based physiotherapy in patients with facial nerve paresis. Journal of the Japanese Physical Therapy Association 2004;7:35 9. Beurskens 2004b Beurskens CHG, Heymans PG. Physiotherapy in patients with facial nerve paresis: description of outcomes. American Journal of Otolaryngology 2004;25(6): Danielidis 1999 Danielidis V, Skevas A, Van Cauwenberge P, Vinck B. A comparative study of age and degree of facial nerve recovery in patients with Bell s palsy. European Archives of Otorhinolaryngology 1999;256(10): De Diego 1999 De Diego JI, Prim MP, Madero R, Gavilán J. Seasonal patterns of idiopathic facial paralysis: a 16-year study. Otolaryngology and Head and Neck Surgery 1999;120(2): Gilden 2004 Gilden DH. Clinical Practice. Bell s palsy. The New England Journal of Medicine 2004;351(13): Gonçalvez 1997 Gonçalvez-Coêlho TD, Pinheiro CN, Ferraz EV, Alonso-Nieto JL. Clusters of Bell s palsy. Arquivos de Neuro-Psiquiatria 1997;55(4): Grogan 2001 Grogan PM, Gronseth GS. Practice parameter: steroids, acyclovir and surgery for Bell s palsy (an evidence based review): report of the Quality Standards Subcommittee of the American Academy of Neurology. Neurology 2001;56(7): Hato 2007 Hato N, Yamada H, Kohno H, Matsumoto S, Honda N, Gyo K, et al.valacyclovir and prednisolone treatment for Bell s palsy: a multicenter, randomized, placebo-controlled study. Otology and Neurotology 2007;28: He 2007 He L, Zhou MK, Zhou D, Wu B, Li N, Kong SY, et al.acupuncture for Bell s palsy. Cochrane Database of Systematic Reviews 2007, Issue 4.[Art. No.: CD DOI: / CD pub3] Holland 2004 Holland NJ, Weiner GM. Recent developments in Bell s palsy. BMJ 2004;329(7465): House 1985 House JW, Brackmann DE. Facial nerve grading system. Otolaryngology--Head and Neck Surgery 1985;93(2): Peitersen 2002 Peitersen E. Bell s Palsy: the spontaneous course of 2500 peripheral facial nerve palsies of different etiologies. Acta Oto-Laryngologica. Supplementum 2002;549:4 30. Quinn 2003 Quinn R, Cramp F. The efficacy of electrotherapy for Bell s palsy: a systematic review. Physical Therapy Reviews 2003;8:

14 Roos 1996 Ross BG, Fradet G, Nedzelski JM. Development of a sensitive clinical facial grading system. Otolaryngology Head and Neck Surgery 1996;114: Salinas 2004 Salinas RA, Alvarez G, Ferreira J. Corticosteroids for Bell s palsy (idiopathic facial paralysis). Cochrane Database of Systematic Reviews 2004, Issue 4.[Art. No.: CD DOI: / CD pub3] Sullivan 2007 Sullivan FM, Swan IR, Donnan PT, Morrison JM, Smith BH, McKinstry B, et al.early treatment with prednisolone or acyclovir in Bell s palsy. New England Journal of Medicine 2007;357(16): Valença 2001 Valença MM, Valença LP, Lima MC. Idiopathic facial paralysis (Bell s palsy): a study of 180 patients [Paralisia facial periférica idiopática de Bell]. Arquivos de Neuro-Psiquiatria 2001;59(3-B): VanSwearingen 1996 VanSwearingen J, Brach J. The Facial Disability Index: reliability and validity of a disability assessment instrument for disorders of the facial neuromuscular system. Physical Therapy 1996;76(12): Indicates the major publication for the study 12

15 C H A R A C T E R I S T I C S O F S T U D I E S Characteristics of included studies [ordered by study ID] Beurskens 2003 Methods Participants Blinding: the assessors of outcomes were unaware of the allocation. Analysis: differences (between the experimental and control group and between pre- and post-tests). Data were collected concerning the level of impairment, disability, and handicap of the patient in pre-test and post-test measures in both the treatment and the control groups. Duration: 3 months of therapy. Follow up: 3 measurement occasions within 1 year: immediately, 3 and 12 months after therapy. Center: Radboud University Nijmegen Medical Center, Nijmegen, The Netherlands and Vrije Universiteit Medical Center, Amsterdam. Design: Randomised clinical trial. N = 50 peripheral facial nerve paresis (34 idiopathic). 2 dropped out in each group. Diagnosis: people with sequelae of facial paralysis, House-Brackmannn IV, for at least 9 months; no nerve or muscle reconstruction; absence of complete, partial, or central facial paralysis; absence of congenital facial paralysis; and sufficient knowledge of the Dutch language. Gender: both sexes (21 males and 29 females), including the participants with other causes of facial palsy. Race: not mentioned. Age: median 44 years (20 to 73, SD 14) including the participants with other causes of facial palsy. Setting: Physiotherapy outpatient department Interventions 1. Exercises (mime therapy) on a individual basis in sessions of 45 minutes, once weekly, over 3 months (10 sessions) and home program of exercises. N = Control group (waiting list). N =18 Outcomes Notes Stiffness of the face. Lip mobility (both lip and pout length). Physical and social index of the Facial Disability Index (VanSwearingen 1996 ) Sunnybrook Facial Grading System. House-Brackmann Facial Grading System. This study description is the pool of three publications by the author about the same population and groups. Risk of bias Item Authors judgement Description Allocation concealment? No C= Inadequate. A coin flip for the first participant and then pairs of patients as they became available 13

16 Flores 1998 Methods Blinding: not done. Analysis: The participants were divided, for purposes of analysis, into those with and those without electromyographic evidence of denervation. Duration: Until functional recovery was achieved according to the May Scale, with evaluations every 14 days. Follow up: not described. Center: Medicina Física y Rehabilitacion Department, Hospital General Regional Num 1, Culiacán, Sinaloa, México. Design: Randomised clinical trial. Participants N = 149 Diagnosis: acute Bell s palsy of onset within 1 to 3 days. EMG 8 days after onset. Excluded other causes of facial paralysis. Gender: both sexes (males 61 and females 88) Race: not mentioned Age: median 33 (3 to 60) years. Setting: clinic. History/Comorbidities: normal glycemia and arterial pressure. Interventions 1. Prednisone (1mg/Kg /day) up 14 days. N = Infrared treatment for 20 min and faradic stimulation (10 to 15 stimulation/min in motor points not described). N = 76 Outcomes Notes Clinical history and May Scale (grade I - complete recovery, II - complete recovery with facial asymmetry with movements between 2 to 6 months, and III - incomplete recovery with asymmetry, synkinesis for more than 6 months). Drop out: 29 people (19.26%) without describing the exact reason for drop out or the groups they were allocated to. Reasons: participants requested another medication or they did not adhere to the treatment. We are waiting for the author s answer about details of the study. Risk of bias Item Authors judgement Description Allocation concealment? Unclear B - Unclear. Allocation not described Manikandan 2007 Methods Blinding: no. Analysis: the authors used Wilcoxon signed-rank test and Mann Whitney U-test to compare the Facial Grading Scale scores within each group. Duration: three months of therapy. Follow up: three months. Until the end of the therapy. Center: Kasturba Hospital, Manipal, Karnataka, India. Design: Randomised clinical trial. 14

17 Manikandan 2007 (Continued) Participants N = 59 participants. Diagnosis: unilateral Bell s palsy with a mean duration of two weeks. Excluded people with diseases of the central nervous system, sensory loss over the face, recurrence of facial paralysis and who were uncooperative during the study. Gender: both sexes (males 24 and females 37). Race: not mentioned. Age: median of 35 years old. Setting: Neurorehabilitation unit. History/Comorbidities: non described Interventions 1. Exercises (facial neuromuscular reeducation)on a individual basis taught to patients, 5 to 10 repetitions, 3 x /day, for 3 months. N = Fixed protocol of electrical stimulation (3 x/day, for six days in 2 weeks. 90 contractions with galvanic current in each muscle plus 10 contractions with faradic current in each facial nerve trunk, intensity until minimal visible contraction)gross facial expression exercises taught to patients for 3 months. N = 30 Participants in both the groups were instructed to use a hand-held mirror during the exercise. Facial massage was given and strapping was applied to the face to maintain the symmetry. Outcomes Notes Facial Grading Scale (facial symmetry at: rest, movement and synkinesis)before and after 3 months. Two patients in group 2 developed mild synkinesis post treatment. One patient from group 1 and two from group 2 dropped out before the completion of the study with reasons stated. Risk of bias Item Authors judgement Description Allocation concealment? Yes A - Adequate. Randomisation using six blocks with 10 in each block. Mosforth 1958 Methods Participants Blinding: none. Analysis: The participants were divided, for purposes of analysis, into those with and those without electromyographic evidence of denervation. Duration: the treatment was given daily until the active contractions returned and then thrice weekly until recovery was virtually complete or the condition seemed stationary (2 to 6 months). Follow up: one year. Center: Department of Electromyography Leeds General Infirmary. Design: controlled randomised trial. N = 86 people with Bell s palsy. Diagnosis: clinically excluding other causes. Complete or partial paralysis of one side of the face, sudden onset. Duration: less than 14 days (mean 5.2). Gender: both sexes males 40 and females

18 Mosforth 1958 (Continued) Race: not mentioned. Age: 37.5 years old (3 to 79 years). Setting: clinic. History/Comorbidities: the groups were comparable at baseline. Interventions 1. Auto-massage of the face plus infrared for 10 min plus interrupted galvanic stimulation of 11 muscles of the face for 3 times of 30 contraction (pulse 100 msec). N = Massage. N = 42 Outcomes Notes Electrical examination. Grade of paralysis estimated visually as a percentage of the function of the normal side. One patient from group 1 and two from group 2 dropped out before the completion of the study with reasons. Risk of bias Item Authors judgement Description Allocation concealment? Yes A - Adequate. A prepared list. Wang 2004 Methods Participants Blinding: no. Analysis: Improvement Index = (Scores After Treatment - Scores Before Treatment)/Scores After Treatment Duration: one month (30 days). Follow up: one month. Until the end of the therapy. Center: Neurology Department of West China Hospital. Design: randomised clinical trial. N = 74 people with Bell s palsy. Diagnosis: diagnosed as facial nerve paralysis by Neurology Department of West China Hospital. Exclusion caused central, traumatic or auditory facial nerve paralysis. Duration: lasting for less than 1 month. Gender: both sexes males 1 and females 0.79 (therapy) and males 1 and females 0.41 (control). Race: Chinese. Age: therapy group mean (SD14.47) years old, and control group (13.46) years. Setting: hospital. History/Comorbidities: not mentioned. Interventions 1. Drug plus physical treatment plus massage plus acupuncture plus functional exercise. N = days - drug treatment and physical treatment days - drug treatment, physical treatment, functional exercise and massage and acupuncture treatment days - physical treatment, functional exercise and massage and acupuncture treatment. 2. Drug plus physical treatment plus massage plus acupuncture. N = days - drug treatment and physical treatment days - physical treatment and massage and acupuncture treatment. 16

19 Wang 2004 (Continued) days - physical treatment and massage and acupuncture treatment. - Drug treatment (cortisone for 30 mg daily in the morning or 10 mg 3 x daily for 7 days, decreased the dosage on the 7th day, and stop on the 14th day; mexobalamin 500 ug 2 x daily; vitamin B2 10 mg). Outcomes Scores of facial muscular function: Potmann Scores (frowning, eyes closing, moving nose, smiling, whistling, and plumping the face, each movement graded 3 scores, adding 2 scores for the impression of quiet state). There was no exact criterion to measure the symptoms. Notes Risk of bias Item Authors judgement Description Allocation concealment? Yes A - Adequate. Randomised numbers by the computer. Wen 2004 Methods Participants Interventions Outcomes Blinding: none. Analysis: each group has patients with three different severities: mild, moderate and severe. The degree of recovery, time to recovery and complications were used to evaluate the results. Follow up: during the treatment = 12 weeks (between 10/2000 and 11/2003). Center: central Hospital of Nanyanz, Manyang, Henan Province, China. Design: controlled randomised trial. N = 145 people with idiopathic facial palsy. Diagnosis: severity based on the function of facial muscles and complaints of patients. Duration: not mentioned. Gender: both sexes males 85 and females 60. Race: not mentioned. Age: 7 to 74 years old (average: 45). Setting: hospital. History/Comorbidities: not mentioned. 1. Conventional therapy plus facial rehabilitation exercises (movements using facial muscles, exercises performed daily under the tutoring of clinicians). N = Conventional therapy only. N = 60. Both groups received the same pharmacological treatment (no information about the dosage that was used). Grade of paralysis estimated visually as a percentage of the function of the normal side. The outcome measures were times when the patient started to recover and completely recovered; the percentage of completely recovered patients within 12 weeks. The measurements took place once a week by clinicians but the results were presented as standard mean differences. No baseline level was indicated. 17

20 Wen 2004 (Continued) Notes Facial muscle synkineses were reported in one case in the mild and one in the moderate group. In the severe patient group, 12 cases of complications reported in the control group and 4 cases in the training group were reported. Risk of bias Item Authors judgement Description Allocation concealment? Unclear B - Unclear. Allocation not mentioned. Characteristics of excluded studies [ordered by study ID] Aleev 1973 Balliet 1982 Bernardes 2004 Beurskens 2004c Brach 1999 Brown 1978 Casler 1990 Coulson 2006 Coulson 2006b Cronin 2003 Dalla-Toffola 2005 Danile 1982 Not a randomised controlled trial. It is a case series. Not a randomised controlled trial. Four people with traumatic facial paralysis. Not a -randomised controlled trial. It is a retrospective study to delineate the contribution of myofunctional exercises during the flaccid phase of the facial paralysis between participants with traumatic and spontaneous paralysis. This is a description of the mime facial exercises. Not a randomised controlled trial. It is a case study that proposed a treatment-based category based on signs and symptoms. Not a randomised controlled trial. It is a case study that described two participants treated with biofeedback in both clinic and home environment. This is a controlled trial about surgery. There were only two participants with idiopathic facial palsy. Not a randomised controlled trial. It is a study of 2 cases following removal of a vestibular schwannoma. Not a randomised controlled trial. This is a retrospective case review. There are others causes of facial palsy including Ramsay Hunt. There were only 3 participants with idiopathic facial palsy. The groups were not comparable at baseline. Twenty-four participants received neuromuscular facial retraining and the other 6 were the control group. Not a randomised controlled trial. This is a retrospective study. Not a randomised controlled trial. Iontophoresis was applied in 50 participants with idiopathic facial palsy without a comparison group. 18

21 (Continued) Dubravica 1996 Goulart 2002 Klingler 1982 Koyama 2005 Lobzin 1989 Manca 1997 Murakami 1993 Nakamura 2003 Romero 1982 Ross 1991 Segal 1995a Segal 1995b Shiau 1995 It was unclear how the groups were divided and if the participants were randomised. The two groups undertook kinesiotherapy plus electrostimulation 5 weeks before the study and it could have interfered with the results. Not a randomised controlled trial. It is a non-systematic review of the literature. This controlled trial is about therapy with cortisone, anti-rheumatics and diuretics to treat facial palsy Not a randomised controlled trial. Not a randomised controlled trial. This is two studies with 32 participants with neuritis and neuropathy of the facial nerve treated with an electromyography feedback device without a comparison group. Not a randomised controlled trial. It is a study of 20 participants with facial paralysis treated with EMG biofeedback. Not a randomised trial. One group of people treated with low reactive-level laser therapy (11) compared with one group treated with stellate ganglion block (26) and another group with a combination of both (15). There were only 10 participants with idiopathic facial palsy. 27 people with complete facial nerve palsy who had no response to electrical stimulation were randomly allocated into 2 groups: 12 treated with training method of biofeedback rehabilitation to prevent synkinesis and 15 as a control without treatment. Not a randomised controlled trial. Biofeedback training was applied in ten participants with at least one-year evolution selected in 957 facial paralyses. Only six of them had idiopathic facial palsy. This study describes a randomised controlled trial with 31 people with long standing facial paresis (minimum of 18 months) but there were only four participants with idiopathic facial palsy. Not a randomised controlled trial. It is a preliminary study with 10 participants. This compared a neuromuscular retraining program (5 participants)to a group with the same treatment plus small movements to limit synkinesis (5 participants). There was one person that did not have idiopathic facial palsy and it is not possible to analyse the data excluding this participant. Not a randomised controlled trial. It was a study of 25 people (5 with idiopathic paralysis)that proposed an exercise program based on home exercises and weekly 50 to 60 minute sessions at the clinic. Reassessment was made at 2.5 month intervals for up to 10 months with the House scale and synkinesis measure. All idiopathic participants changed from House grade 4 to 3 in 5 to 10 months. Not a randomised controlled trial. The assessment was randomised and not the participants. Taverner 1966 This is a randomised clinical trial about adrenocorticotrophic hormone injections. Zhao 2005 This controlled trial is about stellate ganglion block and acupuncture. * EMG = Electromyography 19

22 Characteristics of studies awaiting assessment [ordered by study ID] Diao 2002 Methods Participants Interventions Outcomes Notes Guo 2006 Methods Participants Interventions Outcomes Notes Li 2005 Methods Participants Interventions Outcomes Notes Randomised design. Sample size = 94 (withdrawals: unclear). Experimental Group: 48 acupuncture. Control Group: 46 manipulation. Treatment follow up: after the fourth treatment session.treatment duration: 7 x 4 days. Inclusion: participants with Bell s palsy defined according to clinical diagnostic criteria of all degrees of severity. Aged from 6 to 65, mean age: unclear.male 43, female 51. Experimental group: treatment with acupuncture, five days per week with two rest days. Control group: treatment with manipulation, five days per week. Size of needles: unclear.total number of sites: 11. Length of application: 20 minutes. Length of session: 1 week.total number of treatment sessions: 4. Cured (disappearance of all signs and symptoms, the facial symmetry and the function of mimetic muscle were fully restored after treatment). Markedly effective (the facial symmetry was normal in repose, however, during movement, low-grade paralysis persisted after treatment). Improved (the facial symmetry was improved, however, during movement, paralysis persisted after treatment). No effect (signs and symptoms unchanged after treatment). Experimental group: Cured: 30; Markedly effective:12; Improved:6; No effect:0. Control group: Cured: 29; Markedly effective: 13; Improved: 4; No effect:0. These data were extracted from He

23 Pan 2004 Methods Participants Interventions Outcomes Notes Qu 2005 Methods Participants Interventions Outcomes Notes Wang 2004b Methods Participants Interventions Outcomes Notes Yang 2001 Methods Participants Interventions Outcomes Notes 21

24 Zhang 2005 Methods Participants Interventions Outcomes Notes 22

25 D A T A A N D A N A L Y S E S Comparison 1. ELECTROSTIMULATION VERSUS CONTROL Outcome or subgroup title No. of studies No. of participants Statistical method Effect size 1 Incomplete recovery after 6 and 1 Risk Ratio (M-H, Random, 95% CI) Totals not selected 12 months months 1 Risk Ratio (M-H, Random, 95% CI) Not estimable months 1 Risk Ratio (M-H, Random, 95% CI) Not estimable 2 Mean Facial Grading Scale after 1 Mean Difference (IV, Fixed, 95% CI) Totals not selected 3 months 2.1 Rest score 1 Mean Difference (IV, Fixed, 95% CI) Not estimable 2.2 Movement score 1 Mean Difference (IV, Fixed, 95% CI) Not estimable 2.3 Total score 1 Mean Difference (IV, Fixed, 95% CI) Not estimable Comparison 2. ELECTROSTIMULATION VERSUS PREDNISONE Outcome or subgroup title No. of studies No. of participants Statistical method Effect size 1 Incomplete recovery after six months (all participants) 2 Incomplete recovery six months according severity 2.1 Infrachordal lesion (mild cases) 2.2 Suprachordal lesion (severe cases) 3 Mean time to complete recovery (in days) 3.1 Infrachordal lesion (mld cases) 3.2 Suprachordal lesion (severe cases) 1 Risk Ratio (M-H, Fixed, 95% CI) Totals not selected 1 Risk Ratio (M-H, Fixed, 95% CI) Totals not selected 1 Risk Ratio (M-H, Fixed, 95% CI) Not estimable 1 Risk Ratio (M-H, Fixed, 95% CI) Not estimable Mean Difference (IV, Fixed, 95% CI) [-13.99, -2.77] Mean Difference (IV, Fixed, 95% CI) [-13.13, -1.71] 1 47 Mean Difference (IV, Fixed, 95% CI) [-63.40, ] 23

26 Comparison 3. EXERCISE VERSUS WAITING LIST Outcome or subgroup title No. of studies No. of participants Statistical method Effect size 1 Recovery on Facial Grading 1 Mean Difference (IV, Fixed, 95% CI) Totals not selected Scale (Sunnybrook scale) 2 Recovery on Facial Disability 1 Mean Difference (IV, Fixed, 95% CI) Totals not selected Index-physical 2.1 FDI-physical 1 Mean Difference (IV, Fixed, 95% CI) Not estimable 2.2 FDI-social 1 Mean Difference (IV, Fixed, 95% CI) Not estimable Comparison 4. EXERCISE VERSUS CONVENTIONAL TREATMENT Outcome or subgroup title No. of studies No. of participants Statistical method Effect size 1 Incomplete recovery three 1 Risk Ratio (M-H, Fixed, 95% CI) Subtotals only months after randomisation 2 Mean time from the beginning 1 90 Mean Difference (IV, Fixed, 95% CI) [-1.01, -0.17] of the recovery (in weeks) 2.1 Participants with mild 1 43 Mean Difference (IV, Fixed, 95% CI) [-0.79, 0.19] paralysis 2.2 Participants with moderate 1 47 Mean Difference (IV, Fixed, 95% CI) [-2.22, -0.58] paralysis 3 Mean time from completion of 1 90 Mean Difference (IV, Fixed, 95% CI) [-1.49, -0.34] recovery (in weeks) 3.1 Participants with mild 1 43 Mean Difference (IV, Fixed, 95% CI) [-1.09, 0.29] paralysis 3.2 Participants with moderate 1 47 Mean Difference (IV, Fixed, 95% CI) [-3.15, -1.05] paralysis 4 Potmann Score 1 Mean Difference (IV, Fixed, 95% CI) Totals not selected 24

27 Analysis 1.1. Comparison 1 ELECTROSTIMULATION VERSUS CONTROL, Outcome 1 Incomplete recovery after 6 and 12 months. Review: Physical therapy for Bells palsy (idiopathic facial paralysis) Comparison: 1 ELECTROSTIMULATION VERSUS CONTROL Outcome: 1 Incomplete recovery after 6 and 12 months Study or subgroup Electrotherapy Massage + IV Risk Ratio Risk Ratio n/n n/n M-H,Random,95% CI M-H,Random,95% CI 1 6 months Mosforth /44 11/ [ 0.68, 2.50 ] 2 12 months Mosforth /44 10/ [ 0.55, 2.36 ] Favours treatment Favours control Analysis 1.2. Comparison 1 ELECTROSTIMULATION VERSUS CONTROL, Outcome 2 Mean Facial Grading Scale after 3 months. Review: Physical therapy for Bells palsy (idiopathic facial paralysis) Comparison: 1 ELECTROSTIMULATION VERSUS CONTROL Outcome: 2 Mean Facial Grading Scale after 3 months Study or subgroup Electrical Stimulation Control Mean Difference Mean Difference N Mean(SD) N Mean(SD) IV,Fixed,95% CI IV,Fixed,95% CI 1 Rest score Manikandan (5) 28 5 (4.5) 0.0 [ -2.49, 2.49 ] 2 Movement score Manikandan (14) 28 6 (16.7) [ 59.93, ] 3 Total score Manikandan (22) (18.9) [ 1.26, ] Favours electro Favours control 25

28 Analysis 2.1. Comparison 2 ELECTROSTIMULATION VERSUS PREDNISONE, Outcome 1 Incomplete recovery after six months (all participants). Review: Physical therapy for Bells palsy (idiopathic facial paralysis) Comparison: 2 ELECTROSTIMULATION VERSUS PREDNISONE Outcome: 1 Incomplete recovery after six months (all participants) Study or subgroup Electrostimulation Prednisone Risk Ratio Risk Ratio n/n n/n M-H,Fixed,95% CI M-H,Fixed,95% CI Flores /77 11/ [ 0.38, 1.88 ] Favours treatment Favours control Analysis 2.2. Comparison 2 ELECTROSTIMULATION VERSUS PREDNISONE, Outcome 2 Incomplete recovery six months according severity. Review: Physical therapy for Bells palsy (idiopathic facial paralysis) Comparison: 2 ELECTROSTIMULATION VERSUS PREDNISONE Outcome: 2 Incomplete recovery six months according severity Study or subgroup Electrostimulation Prednisone Risk Ratio Risk Ratio n/n n/n M-H,Fixed,95% CI M-H,Fixed,95% CI 1 Infrachordal lesion (mild cases) Flores /47 0/ [ 0.0, 0.0 ] 2 Suprachordal lesion (severe cases) Flores /30 10/ [ 0.34, 1.15 ] Favours electro Favours prednisone 26

29 Analysis 2.3. Comparison 2 ELECTROSTIMULATION VERSUS PREDNISONE, Outcome 3 Mean time to complete recovery (in days). Review: Physical therapy for Bells palsy (idiopathic facial paralysis) Comparison: 2 ELECTROSTIMULATION VERSUS PREDNISONE Outcome: 3 Mean time to complete recovery (in days) Study or subgroup Electrostimulation Prednisone Mean Difference Weight Mean Difference 1 Infrachordal lesion (mld cases) N Mean(SD) N Mean(SD) IV,Fixed,95% CI IV,Fixed,95% CI Flores (8.92) (19.35) 96.4 % [ , ] Subtotal (95% CI) % [ , ] Heterogeneity: not applicable Test for overall effect: Z = 2.54 (P = 0.011) 2 Suprachordal lesion (severe cases) Flores (44.53) (52.13) 3.6 % [ , ] Subtotal (95% CI) % [ , ] Heterogeneity: not applicable Test for overall effect: Z = 2.26 (P = 0.024) Total (95% CI) % [ , ] Heterogeneity: Chi 2 = 3.00, df = 1 (P = 0.08); I 2 =67% Test for overall effect: Z = 2.93 (P = ) Test for subgroup differences: Chi 2 = 3.00, df = 1 (P = 0.08), I 2 =67% Favours treatment Favours control Analysis 3.1. Comparison 3 EXERCISE VERSUS WAITING LIST, Outcome 1 Recovery on Facial Grading Scale (Sunnybrook scale). Review: Physical therapy for Bells palsy (idiopathic facial paralysis) Comparison: 3 EXERCISE VERSUS WAITING LIST Outcome: 1 Recovery on Facial Grading Scale (Sunnybrook scale) Study or subgroup Treatment Control Mean Difference Mean Difference N Mean(SD) N Mean(SD) IV,Fixed,95% CI IV,Fixed,95% CI Beurskens (18.2) (16.2) [ 8.76, ] Favours control Favours exercises 27

30 Analysis 3.2. Comparison 3 EXERCISE VERSUS WAITING LIST, Outcome 2 Recovery on Facial Disability Index-physical. Review: Physical therapy for Bells palsy (idiopathic facial paralysis) Comparison: 3 EXERCISE VERSUS WAITING LIST Outcome: 2 Recovery on Facial Disability Index-physical Study or subgroup Exercises Control Mean Difference Mean Difference N Mean(SD) N Mean(SD) IV,Fixed,95% CI IV,Fixed,95% CI 1 FDI-physical Beurskens (16.8) (17.9) [ -1.37, ] 2 FDI-social Beurskens (12.2) (16.4) [ 4.85, ] Favours control Favours exercises Analysis 4.1. Comparison 4 EXERCISE VERSUS CONVENTIONAL TREATMENT, Outcome 1 Incomplete recovery three months after randomisation. Review: Physical therapy for Bells palsy (idiopathic facial paralysis) Comparison: 4 EXERCISE VERSUS CONVENTIONAL TREATMENT Outcome: 1 Incomplete recovery three months after randomisation Study or subgroup Facial exercise Control Risk Ratio Risk Ratio n/n n/n M-H,Fixed,95% CI M-H,Fixed,95% CI Wen /85 7/ [ 0.21, 1.71 ] Favours treatment Favours control 28

31 Analysis 4.2. Comparison 4 EXERCISE VERSUS CONVENTIONAL TREATMENT, Outcome 2 Mean time from the beginning of the recovery (in weeks). Review: Physical therapy for Bells palsy (idiopathic facial paralysis) Comparison: 4 EXERCISE VERSUS CONVENTIONAL TREATMENT Outcome: 2 Mean time from the beginning of the recovery (in weeks) Study or subgroup Facial exercise Control Mean Difference Weight Mean Difference 1 Participants with mild paralysis N Mean(SD) N Mean(SD) IV,Fixed,95% CI IV,Fixed,95% CI Wen (0.7) (0.9) 73.7 % [ -0.79, 0.19 ] Subtotal (95% CI) % [ -0.79, 0.19 ] Heterogeneity: not applicable Test for overall effect: Z = 1.19 (P = 0.23) 2 Participants with moderate paralysis Wen (1.3) (1.4) 26.3 % [ -2.22, ] Subtotal (95% CI) % [ -2.22, ] Heterogeneity: not applicable Test for overall effect: Z = 3.33 (P = ) Total (95% CI) % [ -1.01, ] Heterogeneity: Chi 2 = 5.04, df = 1 (P = 0.02); I 2 =80% Test for overall effect: Z = 2.73 (P = ) Test for subgroup differences: Chi 2 = 5.04, df = 1 (P = 0.02), I 2 =80% Favours treatment Favours control 29

32 Analysis 4.3. Comparison 4 EXERCISE VERSUS CONVENTIONAL TREATMENT, Outcome 3 Mean time from completion of recovery (in weeks). Review: Physical therapy for Bells palsy (idiopathic facial paralysis) Comparison: 4 EXERCISE VERSUS CONVENTIONAL TREATMENT Outcome: 3 Mean time from completion of recovery (in weeks) Study or subgroup Facial exercise Control Mean Difference Weight Mean Difference 1 Participants with mild paralysis N Mean(SD) N Mean(SD) IV,Fixed,95% CI IV,Fixed,95% CI Wen (1.2) (1.1) 69.7 % [ -1.09, 0.29 ] Subtotal (95% CI) % [ -1.09, 0.29 ] Heterogeneity: not applicable Test for overall effect: Z = 1.14 (P = 0.26) 2 Participants with moderate paralysis Wen (1.8) (1.7) 30.3 % [ -3.15, ] Subtotal (95% CI) % [ -3.15, ] Heterogeneity: not applicable Test for overall effect: Z = 3.93 (P = ) Total (95% CI) % [ -1.49, ] Heterogeneity: Chi 2 = 7.07, df = 1 (P = 0.01); I 2 =86% Test for overall effect: Z = 3.11 (P = ) Test for subgroup differences: Chi 2 = 7.07, df = 1 (P = 0.01), I 2 =86% Favours treatment Favours control Analysis 4.4. Comparison 4 EXERCISE VERSUS CONVENTIONAL TREATMENT, Outcome 4 Potmann Score. Review: Physical therapy for Bells palsy (idiopathic facial paralysis) Comparison: 4 EXERCISE VERSUS CONVENTIONAL TREATMENT Outcome: 4 Potmann Score Study or subgroup Facial exercise Control Mean Difference Mean Difference N Mean(SD) N Mean(SD) IV,Fixed,95% CI IV,Fixed,95% CI Wang (1.77) (3.75) 8.47 [ 7.05, 9.89 ] Favours control Favours treatment 30

33 A P P E N D I C E S Appendix 1. MEDLINE search strategy 1. exp facial nerve diseases/ 2. bell palsy/ 3. facial paralysis/or hemifacial paralysis/ 4. ((bell$ or facial$ or hemifacial$ or unilateral$ or nerve$ or cranial$) adj3 (pals$ or paralys$ or paresi$ or spasm$)).mp. 5. or/ exp exercise movement techniques/or exp physical therapy modalities/ 7. physical therapy.mp. 8. physio$.mp. 9. rehabilitation.mp. 10. exp Rehabilitation/ 11. exercise$ therapy.mp. 12. Physical Fitness/ 13. physical fitness.mp. 14. Motor Activity/ 15. physical activit$.mp. 16. kinesiotherapy.mp. 17. stretch$.mp. 18. strengthen$.mp. 19. Physical Endurance/ 20. endurance.mp. 21. Biofeedback (Psychology) / 22. biofeedback.mp. 23. Electromyography/ 24. electromyography.mp. 25. electromyogram$.mp. 26. short-wave therapy/ or ultrasonic therapy/ 27. (ultrasonic therapy or short wave therapy).mp. 28. Lasers/ 29. laser$.mp. 30. iontophor$.mp. 31. manipulat$.mp 32. Cryotherapy/ 33. cryotherap$.mp. 34. or/ and randomized controlled trial.pt. 37. controlled clinical trial.pt. 38. randomized controlled trials/ 39. random allocation/ 40. double-blind method/ 41. single-blind method/ 42. or/ animals/ not humans/ not clinical trial.pt. 46. exp clinical trials/ 47. (clin$ adj25 trial$).ti,ab. 48. ((singl$ or doubl$ or tripl$ or trebl$) adj25 (blind$ or mask$)).ti,ab. 49. placebos/ 31

34 50. placebo$.ti,ab. 51. random$.ti,ab. 52. research design/ 53. or/ not not comparative study/ 57. exp evaluation studies/ 58. follow up studies/ 59. prospective studies/ 60. (control$ or prospectiv$ or volunteer$).ti,ab. 61. or/ not not (44 or 55) or 55 or and 64 Appendix 2. EMBASE search strategy 1. exp nerve paralysis/ 2. bell palsy/ 3. facial nerve paralysis/or hemifacial spasm/ 4. ((bell$ or facial$ or hemifacial$ or unilateral$ or nerve$ or cranial$) adj3 (pals$ or paralys$ or paresi$ or spasm$)).mp. 5. or/ exp physiotherapy/or exp kinesiotherapy/ 7. physical therapy.mp. 8. physio$.mp. 9. rehabilitation.mp. 10. exp Rehabilitation/ 11. exercise$ therapy.mp. 12. Fitness/ 13. physical fitness.mp. 14. Motor Activity/ 15. physical activit$.mp. 16. kinesiotherapy.mp. 17. stretch$.mp. 18. strengthen$.mp. 19. Physical Stress/ 20. endurance.mp. 21. psychophysiology/ 22. biofeedback.mp. 23. Electromyography/ 24. electromyography.mp. 25. electromyogram$.mp. 26. diathermy/or ultrasound therapy/ 27. (ultrasonic therapy or short wave therapy).mp. 28. Laser/ 29. laser$.mp. 30. iontophor$.mp. 31. manipulat$.mp. 32. Cryotherapy/ 33. cryotherap$.mp. 32

35 34. or/ and Randomized Controlled Trial/ 37. Clinical Trial/ 38. Multicenter Study/ 39. Controlled Study/ 40. Crossover Procedure/ 41. Double Blind Procedure/ 42. Single Blind Procedure/ 43. exp RANDOMIZATION/ 44. Major Clinical Study/ 45. PLACEBO/ 46. Meta Analysis/ 47. phase 2 clinical trial/ or phase 3 clinical trial/ or phase 4 clinical trial/ 48. (clin$ adj25 trial$).tw. 49. ((singl$ or doubl$ or tripl$ or trebl$) adj25 (blind$ or mask$)).tw. 50. placebo$.tw. 51. random$.tw. 52. control$.tw. 53. (meta?analys$ or systematic review$).tw. 54. (cross?over or factorial or sham? or dummy).tw. 55. ABAB design$.tw. 56. or/ human/ 58. nonhuman/ or not and or and 62 Appendix 3. EBSCOhost CINHAL search strategy S45 S44 and S26 S44 S43 or S42 or S41 or S40 or S39 or S38 or S37 or S36 or S35 or S34 or S33 or S32 or S31 or S30 or S29 or S28 or S27 S43 TI random* or AB random* S42 ( TI (cross?over or placebo* or control* or factorial or sham? or dummy) ) or ( AB (cross?over or placebo* or control* or factorial or sham? or dummy) ) S41 ( TI (clin* or intervention* or compar* or experiment* or preventive or therapeutic) or AB (clin* or intervention* or compar* or experiment* or preventive or therapeutic) ) and ( TI (trial*) or AB (trial*) ) S40 ( TI (meta?analys* or systematic review*) ) or ( AB (meta?analys* or systematic review*) ) S39 ( TI (single* or doubl* or tripl* or trebl*) or AB (single* or doubl* or tripl* or trebl*) ) and ( TI (blind* or mask*) or AB (blind* or mask*) ) S38 ARAB design* S37 PT clinical trial or PT systematic review S36 (MH Factorial Design ) S35 (MH Concurrent Prospective Studies ) or (MH Prospective Studies ) S34 (MH Meta Analysis ) S33 (MH Solomon Four-Group Design ) or (MH Static Group Comparison ) S32 (MH Quasi-Experimental Studies ) S31 (MH Placebos ) S30 (MH Double-Blind Studies ) or (MH Triple-Blind Studies ) 33

36 S29 (MH Clinical Trials+ ) S28 (MH Crossover Design ) S27 (MH Random Assignment ) or (MH Random Sample ) or (MH Simple Random Sample ) or (MH Stratified Random Sample ) or (MH Systematic Random Sample ) S26 S25 and S3 S25 S24 or S23 or S22 or S21 or S20 or S19 or S18 or S17 or S16 or S15 or S14 or S13 or S12 or S11 or S10 or S9 or S8 or S7 or S6 or S5 or S4 S24 cryotherap* S23 (MH Cryotherapy ) S22 manipulat* S21 iontophor* S20 (MH Lasers ) or laser* S19 (MH Ultrasonic Therapy ) or ultrasonic therapy or short-wave therapy S18 electromyogram* S17 (MH Electromyography ) or electromyography S16 (MH Biofeedback ) or biofeedback S15 enduranc* S14 (MH Physical Endurance ) S13 strengthen* S12 stretch* S11 kinesiotherapy S10 physical activit* S9 (MH Motor Activity ) S8 (MH Physical Fitness ) or physical fitness S7 exercise therap* S6 (MH Rehabilitation+ ) or rehabilitation S5 physical therap* or physio* S4 ((MH Physical Therapy+ )) or (MH Therapeutic Exercise+ ) S3 S2 or S1 S2 (bell* or facial* or hemifacial* or unilateral* or nerv* or cranial*) and (pals* or parayls* or paresi* or spasm*) S1 ((MH Bell Palsy ) or (MH Facial Nerve Diseases+ )) or (MH Facial Paralysis ) Appendix 4. LILACS search strategy ((Pt ENSAIO CONTROLADO ALEATORIO OR Pt ENSAIO CLINICO CONTROLADO OR Mh ENSAIOS CONTROLA- DOS ALEATORIOS OR Mh DISTRIBUICAO ALEATORIA OR Mh MÉTODO DUPLO-CEGO OR Mh MÉTODO SIM- PLES-CEGO) AND NOT (Ct ANIMAIS AND NOT (Ct HUMANO AND Ct ANIMAIS)) OR (Pt ENSAIO CLÍNICO OR Ex E $) OR (Tw clin$ AND (Tw trial$ OR Tw ensa$ OR Tw estud$ OR Tw experim$ OR Tw investiga$)) OR ((Tw singl$ OR Tw simple$ OR Tw doubl$ OR Tw doble$ OR Tw duplo$ OR Tw trebl$ OR Tw trip$) AND (Tw blind$ OR Tw cego$ OR Tw ciego$ OR Tw mask$ OR Tw mascar$)) OR Mh PLACEBOS OR Tw placebo$ OR (Tw random$ OR Tw randon$ OR Tw casual$ OR Tw acaso$ OR Tw azar OR Tw aleator$) OR (Mh PROJETOS DE PESQUISA) AND NOT (Ct ANIMAIS AND NOT (Ct HUMANO AND Ct ANIMAIS)) OR (Ct ESTUDO COMPARATIVO OR Ex E05.337$ OR Mh SEGUIMENTOS OR Mh ESTUDOS PROSPECTIVOS OR Tw control$ OR Tw prospectiv$ OR Tw volunt$ OR Tw volunteer$) AND NOT (Ct ANIMAIS AND NOT (Ct HUMANO AND Ct ANIMAIS))) [Palavras] and PARALYSIS$ or (INFLAMMATORY MONONEU- ROPATH$) or (PERIPHERAL NERVOUS SYSTEM DISEASE$) or (HERPES ZOSTER) or NEURALGIA$ or HERPETIC or POSTHERP$ or POST-HERP$ or (DIABETIC NEUROPATH$) or NEURITIS$ or (PERIPHERAL NERVES$) or (NERVE TRAUMA) or (NERVE GRAFT) or (NERVE COMPRESSION SYNDROME$) or (ENTRAPMENT NEUROPATH$) or (BELL PALSY) or (FACIAL PARALYSIS) or (CRANIAL NERVE PALSY) or (FACIAL NERVE INJUR$) or (FACIAL NERVE DISEASE) [Palavras] and (PHYSICAL THERAPY) or PHYSIOTHERAPY or REHABILITATION or (EXERCISE THERAPY) or (PHYSICAL EDUCATION TRAINING) or (PHYSICAL FITNESS) or (PHYSICAL ACTIVIT$) or KINESIOTHERAP$ or STRETCHING or STRENGTHENING or ENDURANCE or BIOFEEDBACK or ELECTROMYOGRAPHY or ELECTROMYOGRAPHY$ or 34

37 ELECTROMYOGRAM$ or BIOFEEDBACK or ULTRASON$ or LASER$ or (SHORT WAVE THERAPY) or IONTOPHOR$ or MANIPULAT$ or CRYOTHERAPY [Palavras] Appendix 5. PEDro search strategy Abstract & Titles: facial palsy Therapy: no selection Body part: no selection Subdiscipline: no selection Method: no selection When searching: match all search terms (AND) H I S T O R Y Protocol first published: Issue 3, 2006 Review first published: Issue 3, March 2008 Amended Converted to new review format. C O N T R I B U T I O N S O F A U T H O R S LJT suggested the review, reviewed the literature wrote the primary version of the protocol and the review. VPV revised the protocol, the language and the search strategy. BGOS supervised the process of the protocol development and made contributions about the methodology. GFP gave specialist contributions. All authors approved the final text of the review. D E C L A R A T I O N S O F None known. I N T E R E S T I N D E X T E R M S 35

38 Medical Subject Headings (MeSH) Physical Therapy Modalities; Bell Palsy [ therapy]; Electric Stimulation Therapy [methods]; Exercise Therapy [methods]; Facial Muscles; Massage [methods]; Randomized Controlled Trials as Topic MeSH check words Humans 36