Study	Reason for exclusion
Abashev 1997	Location: Russia Excluded because we were unable to determine if this is a randomized controlled trial. The trial report says 5 study groups were formed using the results of the therapy, but continues to say that the 5 groups were different because they received different therapy: (1) antibacterial; (2) antibacterial plus LLLT of projections to surfaces of the body close to the internal organs affected by tuberculosis (ie, anatomical projections of affected areas); (3) antibacterial plus ultrasound therapy; (4) antibacterial plus LLLT on anatomical projections of affected organs plus ultrasound therapy; and (5) antibacterial plus LLLT on acupuncture points plus ultrasound therapy. In correspondence received from the author, the authors stated that the "groups were formed by the method of random sample". However, the trial author did not explicitly state that this was a prospective study, and if participants were randomly assigned to groups, what method was used for randomization; further attempts to contact the trial author were unsuccessful
Agayev 2004	Location: Azerbaijan Study not randomized
Lovacheva 2006	Location: Russia The method of randomization was not reported. After being contacted, the leading author explained that there was no systematic application of true randomization procedure for allocation of treatments.
Maliev 1991a	Also see Shesterina 1991 and Maliev 2001b Location: Russia Excluded because we were unable to determine if this is a randomized controlled trial. This study was first published as a controlled trial or comparison of patients selected from the current practice. In Maliev 2001b, the same study was described as 'randomized'. The authors did not mention the method of randomization, concealment of allocation, or blinding
Maliev 2001b	Also see Maliev 1991a and Shesterina 1991 Location: Russia Excluded because there is a discrepancy in the number of participants included in the study. The same group of authors wrote this study report as for Maliev 1991a and Shesterina 1991. The study description is very similar to Maliev 1991a, although the immunoglobulin results are presented more extensively. While Maliev 1991a mentioned that two participants dropped out because of poor adherence, Shesterina 1991 reported one dropout and Maliev 2001b also reports one dropout. Despite the different number of participants in Maliev 1991a and Shesterina 1991, the results were described with identical numbers (see Shesterina 1991)
Shesterina 1991	Also see Maliev 1991a and Maliev 2001b Location: Russia Excluded because there is a discrepancy in the number of participants included in the study. The same group of authors wrote this study report as for Maliev 1991a. The study design was the same, but number of participants was only 56, not 102. While Maliev 1991a mentioned that 2 participants dropped out because of low adherence, this study (Shesterina 1991) did not report any dropouts. In a further publication of the data of the same 56 participants (Maliev 1991b ‐ Additional references), 1 participant was reported to have dropped out. In the latest publication (Maliev 2001a), 5 participants in the intervention group and 5 participants in the control group are reported to have dropped out. Despite different number of participants in two reports (Maliev 1991a, Shesterina 1991), the outcomes were described with identical numbers as shown below: Maliev 1991a: 1. Catarrhal endobronchitis "clinically cured" in 25/25 participants during 0.99+‐0.1 months of treatment in the intervention group and in 18/28 participants during 2.58+‐0.23 months in the control group. 2. Purulent endobronchitis relying on the bronchoscopy was cured in 25/26 participants (1 dropout/17 participants) during 1.13+‐0.11 months in the intervention group and 13/22 participants during 3.43+‐0.51 months in the control group Shesterina 1991: 1. Catarrhal endobronchitis "clinically cured" in 13/13 participants during 0.99+‐0.09 months of treatment in the intervention group and 9/14 participants during 2.58+‐0.23 months in the control group. 2. Purulent endobronchitis relying on the bronchoscopy was cured in 15/16 participants (1 dropout/17 participants) during 1.13+‐0.11 months in the intervention group and 7/12 participants during 3.43+‐0.51 months in the control group Repeated attempts to contact investigators to clear the problems with publications and study design were unsuccessful. In the light of these discrepancies we are unable to include this study
Topolnitskii 1992	Location: Russia Excluded because the intervention and control groups did not receive the same baseline treatments: the control group received "pathogenetic" interventions (methyluracil, insulin, sodium thiosulphate) and no LLLT; the intervention group received no "pathogenetic" interventions, but did receive LLLT. The report in the full text dissertation says participants were distributed into two groups using a random numbers table. Concealment of allocation and blinding were not mentioned. Of the 25 participants in the intervention group, 6 were excluded because of poor adherence and discharged from the hospital between 1 to 3 months of therapy. No participants were excluded from the control group. The outcomes presented for the intervention group were calculated only for the participants that remained in the study. The trial author did not respond to our requests for further information

LLLT: low level laser therapy.