The inter-rater agreement for study screening between investigators was kappa = 0.79. In total, 308 articles were searched in the databases. After duplicates were removed, 135 articles were retained for title and abstract identification. The full texts of 17 articles were read carefully and further assessed in strict accordance with the eligible criteria. Three RCTs were excluded; two RCTs did not include dialysis treatment [27, 28], and one was still ongoing . Eventually, five RCTs were included for systematic assessment (Fig. 1). Screening the reference lists did not reveal additional suitable articles in accordance with the inclusion criteria.
Characteristics of the included studies
Table 1 shows details of the summarized characteristics of included studies. 5 included RCTs [23, 24, 30-32] were of parallel design. Fang , Li  and Ma  recruited HD patients, Zhang and colleagues  recruited PD patients, and Wehmeyer and colleagues  recruited HD and PD patients. The HD treatment time established by investigators of three studies was on average 4 h of HD 3 times a week [23, 24, 32]. In the study by Zhang  and Wehmeyer , the HD or PD treatment time was not mentioned. All of included study were at least 1 month follow-ups. The investigators of two studies included both a 3-month and 6-month follow-up periods [24, 30]. Four studies reported hs-CRP levels [23, 24, 31, 32], two reported IL-6 and Alb levels [24, 30], and only one reported serum TNF-a level and lipid metabolism markers . The data of hs-CRP levels are available in 4 studies and were collected for the meta-analysis. Five studies provided approximately 353 subjects. However, in the study by Fang and colleagues, two patients were lost to 6-month follow-up owing to time conflicts in the intervention group; two patient was lost to 6-week and 3-month follow-up in the control group, respectively .
All studies described similar NSPT based on OHI and SRP without surgical procedures. One study combined SRP with local minocycline administered to all sites with PD >5 mm at baseline, 3- and 6- month follow up visits . Patients in the studies conducted by Zhang , Li , Ma  and Wehmeyer  did not receive prophylaxis within follow up periods. However, Fang and colleagues carried out supragingival prophylaxis at 3 months .
Risk of bias
Table 2 presents the methodological and quality of the trials included in the review. Because most of the included studies did not report random sequence generation, and none of the studies reported blinding of outcome assessment of practitioners. All considered studies were judged to exhibit a moderate-risk bias.
The descriptive synthesis of the included studies is displayed in Table 1. Only one study focused on TNF-a and lipid metabolism , thus, MAs for these markers were not possible. The levels of TNF-a and lipid metabolism markers showed no significant difference at most points. However, the level of HDL-C was significantly decreased at 6 weeks after NSPT in the HD and/or PD patients. None of the included studies reported the occurrence of adverse effects related to NSPT in the HD and/or PD patients.
Four studies [23, 24, 31, 32] reported the levels of hs-CRP in dialysis patients after non-surgical periodontal treatment. The duration of the intervention of 4 studies varied from 4 weeks to 6 months, so we performed subgroup analysis of the follow-up periods less than or equal to 2 months, 3 months and 6 months. The results of this subgroup analysis showed that the level of serum hs-CRP was significantly decreased at less than or equal to 2 months (SMD: -1.53; 95 % CI: -2.95 to -0.11), 3 months (SMD: -0.27; 95 % CI: -0.47 to -0.07) and 6 months (SMD: -0.36; 95 % CI: -0.55 to -0.17) in dialysis after NSPT, compared with untreated periodontitis patients receiving dialysis (Fig. 2), suggesting the hs-CRP levels were significantly decreased in dialysis patients with periodontitis following NSPT. The heterogeneity observed among the studies was high, so a random-effect model was used.
IL-6 and Alb
Two studies [24, 30] reported IL-6 and Alb levels after NSPT for HD and/or PD patients with CP. The subgroup analysis revealed that after NSPT, the levels of IL-6 in dialysis patients were decreased at 3 months (SMD: -0.03; 95 % CI: -0.84 to 0.78) and 6 months (SMD: -0.23; 95 % CI: -0.78 to 0.33), while there was no significant difference (Fig. 4); and the subgroup analysis revealed that the levels of Alb in the HD and/or PD patients were increased at 3 months (SMD: 1.54; 95 % CI: -0.29 to 3.37) and 6 months (SMD: 1.36; 95 % CI: -0.22 to 2.94). However, no significant difference was found (Fig. 5). Considering that the heterogeneity between the studies was high, we selected a random effect model.