Strategies to increase the coverage of influenza and pneumonia vaccination in older adults: a systematic review and network meta-analysis

Abstract

Background

It is urgent to implement interventions to increase vaccination rates of influenza/pneumonia vaccines in older adults, yet the effectiveness of different intervention strategies has not been thoroughly evaluated.

Objective

We aimed to assess the effectiveness of intervention strategies for increasing the coverage of influenza/pneumonia vaccination in older adults.

Methods

PubMed, Web of Science, Cochrane Library, Embase, China Biology Medicine disc, China National Knowledge Infrastructure and Wanfang were searched from 1 January 2000 to 1 October 2022. RCTs that assessed any intervention strategies for increasing influenza/pneumonia vaccination coverage or willingness in older adults were included. A series of random-effects network meta-analysis was conducted by using frequentist frameworks.

Results

Twenty-two RCTs involving 385,182 older participants were eligible for further analysis. Eight types of intervention strategies were evaluated. Compared with routine notification, health education (odds ratio [OR], 1.85 [95%CI, 1.19 to 2.88]), centralised reminder (OR, 1.63 [95%CI, 1.07 to 2.47]), health education + onsite vaccination (OR, 2.89 [95%CI, 1.30 to 6.39]), and health education + centralised reminder + onsite vaccination (OR, 20.76 [95%CI, 7.33 to 58.74]) could effectively improve the vaccination rate. The evidence grade was low or very low due to the substantial heterogeneity among studies.

Conclusions

Our findings suggest that health education + centralised reminder + onsite vaccination may potentially be an effective strategy regardless of cost, but the evidence level was low. More rigorous trials are needed to identify the association between strategies and vaccination rates among older adults and to integrate such evidence into clinical care to improve vaccination rates.

Key Points

• The best intervention strategy to promote influenza/pneumonia vaccination in older adults remains unclear.

• This network meta-analysis that included 22 randomised controlled trials with 8 types of intervention strategies, the proportion achieving a significantly higher odds ratio was 1.85 with health education, 2.89 with health education + onsite vaccination, and 20.76 with health education + centralised reminder + onsite vaccination compared with routine notification.

• Without considering the economic cost, health education + centralised reminder + onsite vaccination may potentially be an effective intervention strategy.

Introduction

Influenza and pneumonia are two common respiratory infectious diseases that impose a heavy disease burden on older adults, which are associated with a high risk of hospitalization mortality among older adults, and increased healthcare costs [1, 2]. From 2004 to 2012, a mean of 2,235 pneumonia hospitalization deaths in older adults were recorded in Australia, with a case fatality rate of 6.1% [3]. It was estimated that there were 35 million flu-related illnesses, 16 million flu-related medical visits, 380,000 flu-related hospitalisations and 20,000 flu-related deaths in the 2019 to 2020 flu season in the United States [4]. Of these, people 65 years and older accounted for 45% of hospitalizations and 58.7% of deaths [4]. In China, the incidence of community-acquired pneumonia increased rapidly among older adults, ranging from 7.8 per 1,000 person-years among people over 60 years to 14.9 in those over 80 [5]. Overall mean length of hospital stay was 12.1 days, the in-hospital mortality rate was 11.26%, and mean costs were €8,301 in hospitalized older with community-acquired pneumonia in the Netherlands [6].

Influenza/pneumonia vaccination is one of the most effective ways to protect older adults against flu or pneumonia, preventing potentially serious complications, and reducing associated morbidity and mortality [7, 8]. Influenza vaccination coverage varies considerably between countries although World Health Assembly resolution 58 urging Member States to attain influenza vaccination coverage of 75% among older adults by 2010 [9]. Influenza vaccination rates in countries in the Americas varied from 6.8% to 100%, while in Europe they varied from below 1% to over 75% among older people [10, 11]. Studies in recent years have also shown that vaccination rates vary from country to country or region to region. In Switzerland, influenza vaccination coverage among people 65 years and older decreased from 47.8% in 2007 to 36.2% in 2017 [12]. In 2014, 61.3% of adults aged 65 years and older of the United States received their recommended pneumococcal vaccines [13]. However, the coverage of the 23-valent pneumococcal polysaccharide vaccine in older adults ranged from 0.17% to 0.69%, and the overall coverage was higher in urban areas (3.70%) than in rural (3.34%) and suburban areas (2.16%) in 2017 in Hangzhou, China [14]. In fact, countries around the world have taken active interventions to increase influenza/pneumonia vaccination rates in older adults, but the effectiveness of the implementations remains uncertain [15–17].

Previous systematic reviews on the effects of influenza or pneumonia vaccination interventions have some limitations, including outdated searches, poorly targeted studies on effectiveness of vaccination interventions (large varieties of vaccines, wide range of populations), inadequate risk of bias assessment, lack of statistical pooling, and failure to evaluate the quality of evidence [18–20]. Still there is no recent systematic review on how to improve influenza/pneumonia vaccination coverage in older adults. Given the current global epidemic of COVID-19, it is a matter of concern to improve the vaccination rate of influenza/pneumonia vaccines for older adults who are susceptible to respiratory diseases and are prone to adverse consequences. In fact, strategies to increase immunsation coverage among older adults were widely used both in the past and today, including health education, centralised reminder and so on [21–23]. However, the comparative effectiveness between different interventions or strategies remains unknown. Therefore, we conducted a network meta-analysis (NMA) to systematically analyse and compare the effectiveness of different interventions to identify the best strategies to promote influenza/pneumonia vaccination among older adults. We conducted subgroup analyses by vaccine types, sources of subjects, and regional economic development levels. In this way, we hope to summarise intervention strategies to improve influenza/pneumonia vaccination rates among older adults, as well as recommendations for interventions in different settings.

Methods

We registered this review in the International prospective register of systematic reviews (identifier number: CRD42021284085) and followed the Preferred Reporting Items for Systematic Reviews and Meta-analyses guidelines for systematic reviews and meta-analyses [24].

Eligibility criteria

We included randomized controlled trials (RCTs) that compared any interventions or strategies to improve the vaccination rate or willingness of influenza/pneumonia in older adults (aged 60 years and above). Studies included were limited to those published in English or Chinese since 2000. We excluded studies that did not meet criteria by browsing titles, abstracts and full articles. If the study was published more than once, we selected the newest article which has the most complete data or the largest sample size.

Search strategy

We searched the following databases with the assistance from academic librarians: the Cochrane Library, MEDLINE (via PubMed), EMBASE, Web of Science, China Biology Medicine disc, China National Knowledge Infrastructure and Wanfang database from 1 January 2000 to 1 October 2022. The search strategies were provided in eTable 1 in Supplement. We also searched Google Scholar, the preprint servers medRxiv (https://www.medrxiv.org/) and bioRxiv (https://www.biorxiv.org/) for grey literature, and reviewed the reference lists of included studies and previous systematic reviews.

Study selection and data extraction

Two reviewers (X Ma and J Wang) independently screened the potentially eligible titles/abstracts of identified citations by using Endnote, and then selected the final included studies by browsing the full texts. Reviewers resolved disagreements by discussion, or with the help from an adjudicator (P Du).

A pre-designed, pilot-tested data extraction form was used to extract information, including [1] general information (title, author, year of publication, country, study design, setting, start time and follow-up); [2] patient characteristics (age, sex, type of vaccination, source of study subjects, payment); [3] interventions (see below); [4] outcome measures; and [5] risk of bias assessment. The primary outcome was the vaccination rate, which was defined as the rate of influenza or pneumonia vaccination in people aged 60/65 years and older after receiving any intervention. The secondary outcomes were vaccination willingness and costs of the interventions.

Risk of bias assessment and quality of evidence

The Cochrane Risk of Bias 2 (ROB 2) tool was used to assess the quality of the RCTs [25]. The Grading of Recommendations Assessment, Development and Evaluation (GRADE) was used to appraise the quality of evidence from the NMA [26]. The details are described in the eMethods in the Supplement.

Statistical analysis

We used Sankey diagrams for the basic information of the included studies and drew the network meta-analysis map of each intervention using Stata software version 15.1. We used odds ratio (OR) and 95% confidence interval (95% CI) for the dichotomous outcomes. Initially, we performed a conventional pairwise meta-analysis using a DerSimonian–Laird random-effects model and then performed a frequentist NMA using ‘netmeta’ package of R software version 3.6 [27]. The interventions are ranked by the P-score of netrank which takes the value from 0 to 1 with higher scores indicating more effective interventions [28]. Results of NMA were aggregated in league tables using the ‘netleague’ package with Routine notification (RN) as the reference indicator. We assessed the heterogeneity of included studies for each direct comparison with visual inspection of forest plots. Random-effects model was used to obtain pooled ORs (with associated CIs) for the outcome, with the I² statistic used to quantify heterogeneity. We used both funnel plots and Egger’s test to assess potential publication bias when at least 10 studies reported the outcome.

We used the ‘design-by-treatment’ model (global test) to assess the coherence assumption for each network [29]. The Z-test based back calculation method was mainly used to evaluate the inconsistency between the direct and indirect comparison results [30]. The following subgroup analyses were conducted to explore the vaccination coverage rates in different settings: (i) influenza vs pneumonia vaccination; (ii) patients (recruited from clinics, hospitals, community health service centers, and other medical institutions) vs general population (community population from non-medical institutions such as streets, churches, and communities); and (iii) developed vs developing countries/regions [31]. A two-sided P < 0.05 was considered statistically significant.

Results

Study characteristics

Of 6,433 citations, 23 studies were included (of which 5 were cluster RCTs) (Figure 1). The characteristics of the included studies were provided in eTable 2. Twenty-three eligible studies including 389,182 older adults were published from 2002 to 2022. Nine studies were conducted in China [16, 22, 32–38], six in the United States [39–44], four in the United Kingdom [45–48], two in Singapore [17, 49], one in Thailand [50] and Spain [51], respectively. The participants of 17 studies were patient populations and 6 studies were general community populations. Eleven studies reported interventions to promote influenza vaccine, 8 studies for pneumonia vaccine and 4 studies for both vaccines. The median duration of interventions was 5 months, ranging from 12 days to 17 months [34]. Among these 23 studies, 19 (82.6%) reported the effect of interventions on improving influenza or pneumonia vaccination rates among older adults, and 4 (17.4%) reported the effect on vaccination willingness. Twelve studies provided free vaccination, but participants paid the vaccination in 4 studies, and the remaining 7 did not report the payment of vaccination (Figure 2).

Intervention strategies

Based on the summary of all the included studies, the intervention strategies were divided into eight types according to the different implementation content and methods: control intervention (Routine notification); four single-type interventions (Health education, Centralised reminder, Health prescriptions based on questionnaire, Financial incentive); three mixed interventions (Nursing interventions + Onsite vaccination, Health education + Onsite vaccination, Health education + Centralised reminder + Onsite vaccination). The details of the interventions are presented in Table 1.

Risk of bias

Of the 23 RCTs, 14 (60.9%) studies were high risk [16, 17, 22, 32–35, 36, 38, 38, 43, 47, 50, 51], 8 (34.8%) were some concerns [39, 41, 42, 44–46, 48, 49] and 1 (4.3%) was low risk [40] (eFigure 1, Supplementary data are available in Age and Ageing online). The details of RoB for each RCT are provided in eTable 3, Supplementary data are available in Age and Ageing online.

Network meta-analysis

A total of 22 of the 23 studies were included in the NMA, while only one study using financial incentives as an intervention was not suitable for inclusion in the NMA because it could not form a network link with other interventions [49]. Figure 3 presents NMA maps of influenza/pneumonia vaccination rates. The rankings of intervention strategies in order of effectiveness are presented in Table 2.

Total vaccination rate

Nineteen studies reported influenza/pneumonia vaccination rates involving 293,322 participants, seven intervention strategies and seven direct comparisons with a closed-loop (Figure 3A). Regardless of the type of interventions, the conventional meta-analysis showed that, compared to control group, the interventions significantly increased vaccination rates: OR 1.25 [95% CI, 1.23 to 1.28] (eFigure 2, Supplementary data are available in Age and Ageing online). Compared with RN, low- or very low-certainty evidence showed that four intervention strategies significantly increased vaccination rates, including HE: OR, 1.89 [95% CI, 1.19 to 2.88], CR: OR, 1.63 [95% CI, 1.07 to 2.47], HE+OV: OR, 2.89 [95% CI, 1.30 to 6.39], HE+CR + OV: OR, 20.76 [95% CI, 7.33 to 58.74], P < 0.001, I² = 98.2%) (Figure 4A). The details of pair-to-pair comparison results of each intervention and the certainty of evidence are shown in eTable 4 and eTable 5, Supplementary data are available in Age and Ageing online, respectively. A comparison of the impact of different intervention strategies on boosting vaccination rates through the NMA found that HE+CR + OV (P-score = 0.9994) was the best strategy, followed by HE+NI (P-score = 0.7174) (Table 2). Consistency test analysis revealed inconsistency (P = 0.084) (eFigure 3, Supplementary data are available in Age and Ageing online). Publication bias assessment found that the funnel plot showed asymmetry (P = 0.026) (eFigure 4, Supplementary data are available in Age and Ageing online).

Subgroup analysis

Influenza vaccination rate

Thirteen studies reported influenza vaccination rates involving 262,962 participants, seven intervention strategies, and six direct comparisons with no closed-loop (Figure 3B). Compared with RN, low- or very low- certainty evidence showed that the intervention strategies significantly increased influenza vaccination rates (HE+OV: OR, 8.67 [95% CI, 1.60 to 46.86], HE+CR + OV: OR, 20.76 [95% CI, 5.31 to 81.09], P < 0.001, I² = 98.9%) (Figure 4B, eTable 5). In the NMA of intervention strategies to increase influenza vaccination rates, HE+CR + OV (P-score = 0.9607) was the best strategy, followed by HE+OV (P-score = 0.8314) (Table 2). Publication bias assessment revealed that the funnel plot showed asymmetry (P = 0.040) (eFigure 3, Supplementary data are available in Age and Ageing online).

Pneumonia vaccination rate

Eleven studies reported pneumonia vaccination rates involving 44,377 participants, five intervention strategies, and five direct comparisons with a closed-loop (Figure 3C). Low- or very low- certainty evidence showed that HE (OR, 1.78 [95% CI, 1.09 to 2.91]) and HE+OV (OR, 2.37 [95% CI, 1.16 to 4.84]) (P < 0.001, I² = 90.9%) significantly improved pneumonia vaccination rates compared with RN (Figure 4C, eTable 5, Supplementary data are available in Age and Ageing online). HE+OV (P-score = 0.8802) was the best strategy to improve pneumonia vaccination rates. The results of the consistency test revealed no inconsistency (P = 0.176) (eFigure 3, Supplementary data are available in Age and Ageing online). Visual examination of the funnel plot indicated that no significant publication bias was present (P = 0.118) (eFigure 4, Supplementary data are available in Age and Ageing online).

General population vaccination rate

Four studies reported the impact of interventions on vaccination rates in general populations involving 22,920 participants, four intervention strategies, and four direct comparisons with a closed-loop (Figure 3D). Low-certainty evidence showed that HE+CR + OV (OR, 20.76 [95% CI, 3.38 to 127.58]) (P = 0.021, I² = 89.5%) significantly improved vaccination rates compared with RN (Figure 4D, eTable 5, Supplementary data are available in Age and Ageing online), and was the best strategy (P-score = 0.9795). The inconsistency test results found no inconsistency (P = 0.317) (eFigure 3, Supplementary data are available in Age and Ageing online).

Patient population vaccination rate

Fifteen studies reported the impact of interventions on vaccination rates in patient populations involving 270,402 participants, five intervention strategies, four direct comparisons with no closed-loop (Figure 3E). Very low-certainty evidence showed that HE (OR, 2.01 [95% CI, 1.23 to 3.27]) and CR (OR, 1.63 [95% CI, 1.07 to 2.46]) (P < 0.001, I² = 98.4%) significantly improved vaccination rates compared with RN (Figure 4E, eTable 5, Supplementary data are available in Age and Ageing online). HE (P-score = 0.7377) was the best strategy (Table 2). Publication bias assessment revealed that the funnel plot showed asymmetry (P = 0.046) (eFigure 4, Supplementary data are available in Age and Ageing online).

Vaccination rate in developed areas

Fourteen studies reported intervention conducted in developed areas involving 191,061 participants, six intervention strategies and five direct comparisons with no closed-loop (Figure 3F). Low- or very low- certainty evidence showed that CR (OR, 1.63 [95% CI, 1.07 to 2.47]) and HE+OV (OR, 4.90 [95% CI, 1.28 to 18.78]) (P < 0.001, I² = 98.6%) significantly improved vaccination rates compared with RN (Figure 4F, eTable 5, Supplementary data are available in Age and Ageing online) which was the best strategy (P-score = 0.9328) (Table 2). Publication bias assessment revealed that the funnel plot showed asymmetry (P = 0.046) (eFigure 4, Supplementary data are available in Age and Ageing online).

Vaccination rate in developing areas

Five studies reported intervention conducted in developing areas involving 102,261 participants, four intervention strategies, and three direct comparisons with no closed-loop (Figure 3G). Low-certainty evidence showed that HE (OR, 3.54 [95% CI, 1.61 to 7.79]) and HE+CR + OV (OR, 20.76 [95% CI, 6.33 to 68.02], P = 0.021, I² = 74%) significantly increased vaccination rates compared to RN (Figure 4G, eTable 5, Supplementary data are available in Age and Ageing online), with HE+CR + OV (P-score = 0.9970) been the best strategy to improve vaccination rates (Table 2).

Vaccination willingness

Four studies reported vaccination willingness involving 6,143 participants, three intervention strategies, two direct comparisons, and no closed-loop (Figure 3H). Moderate-certainty evidence showed that HE (OR, 2.40 [95% CI, 1.85 to 3.11]) and HE+OV (OR, 2.42 [95% CI, 1.77 to 3.31]) (P < 0.001, I² = 22.1%) significantly increased vaccination willingness compared with RN (Figure 4H, eTable 5, Supplementary data are available in Age and Ageing online), with HE+OV been the best strategy (P-score = 0.7574) (Table 2).

Costs of the intervention

Of the 23 included studies, only 6 summarised intervention costs in the results section [23, 39, 40, 44, 45, 48, 49, 50] and 3 mentioned intervention costs in the discussion section [17, 35, 46] (eTable 6, Supplementary data are available in Age and Ageing online).

Discussion

In this NMA, we combined direct and indirect evidence from 22 RCTs to compare various intervention strategies which aim to improve influenza/pneumonia vaccination rates in older adults. In these studies, all of the intervention strategies demonstrated a degree of effectiveness in increasing influenza/pneumonia vaccination rates or increasing willingness to vaccinate among older adults.

Among the single-type interventions, Health education might be the simplest and effective intervention. Previous studies had shown that the main reasons for low influenza/pneumonia vaccination rates among older adults were the lack of knowledge about influenza/pneumonia vaccine and the misconception that vaccination would produce side effects [20, 32]. Compared to routine notification, the researchers implementing HE effectively increased the awareness of older adults about the safety and efficacy of influenza/pneumonia vaccines by distributing informational materials, posting educational posters, and playing promotional videos, which significantly increased vaccination willingness and vaccination rates [37, 42]. Some studies had also shown that most older people attach more importance to the opinions of health care professionals and recognize their authority [38, 51]. Therefore, physicians or community health care professionals should become active advocates and promoters of influenza/pneumonia vaccination. However, the intervention effect of single HE in economically developed areas, as well as general populations was poor [22, 42]. This may be due to the fact that compared to patients, the general population was in better health status, less aware of health risks, and had a lower need for disease prevention services. So, it may be difficult to motivate them to adopt vaccination behaviours through HE alone. The economically developed areas have achieved better vaccine publicity in their daily work, and the population has a higher education level and a better understanding of health knowledge. Therefore, HE alone may have diminished marginal effect and fail to produce significant result, and needs to be combined with other interventions to increase the vaccination rates.

Among other single-type interventions, centralised reminder which regularly reminded older adults of influenza/pneumonia vaccine by phone and letter [44, 45], was associated with effective improvement of vaccination rates, while questionnaires based on the question-behaviour effect [48] were not. Centralised reminder could achieve certain intervention effects in the patient population [40], which might be related to the fact that patients were more concerned about their health status and accept and comply with medical advice. The adoption of financial incentive, while somewhat improving the vaccination rates [49], was not suitable for wider regional roll-out due to its high cost resulting in lower cost–benefit.

Among the three compound-type interventions, health education + onsite vaccination, and health education + centralized reminder + onsite vaccination could possibly improve influenza/pneumonia vaccination rates in older adults, while nursing interventions + onsite vaccination was less effective. According to the Knowledge Attitude and Practice theoretical model [52], the change of health-related behaviour will undergo a complex and long process, and receiving health-related information is the basis for generating behaviour. Therefore, health education is a prerequisite for intervention, and the poor effect of the nursing interventions+onsite vaccination intervention strategy may be related to the lack of health education. More importantly, establishing beliefs and changing attitudes is harder than acquiring knowledge. Previous research had shown that older adults refuse influenza/pneumonia vaccination partly because of a lack of self-perceived risk [36]. When using the nursing interventions, health care staff built a trusting relationship through regular follow-up visits to households, comprehensively assessed the health status of older adults, guided them to raise awareness of health risk, helped them to effectively receive and reinforce information about influenza/pneumonia vaccination, and further elevated knowledge to beliefs, thereby generating motivation and increasing vaccination willingness and rates [46]. The onsite vaccination strategy of setting up additional temporary vaccination sites in health care facilities or older families could reduce vaccination hesitation so that older adults who is generating motivation can be vaccinated immediately, which improve the accessibility and convenience of vaccination services, enhance the effectiveness of health education, and prevent older people from giving up vaccination due to mobility problems or unsuitable timing [35].

The combination of three interventions health education + centralised reminder + onsite vaccination may potentially be a more effective strategy, which has also played an essential role in promoting COVID-19 vaccine acceptance among older adults. A community study conducted in Singapore increased the COVID-19 vaccination rates among older adults from 84.6% to 96.3% by addressing their concerns and facilitating their vaccination appointment through brief phone conversations [53]. But it requires certain organisation and coordination skills, the mobilisation of more resources, as well as better staff and technical support, which may be more challenging for economically disadvantaged areas [34]. Therefore, the implementation of intervention strategies should not only consider the effect, but also take full account of local health policies, economic conditions and other influencing factors.

Older adults are recommended for specific vaccinations in many countries, and were among the first to be vaccinated during the current SARS-CoV-2 pandemic due to their higher risk of severe illness. Moreover, older adults should also be proactively vaccinated and protected against other respiratory viruses, such as respiratory syncytial virus, to ensure healthy aging and maintain quality of life. The focus on vaccine development should be accompanied by an exploration of more effective vaccination strategies to strengthen preventive care and community immunization in the health care system.

Limitations

Our study had five main limitations. First, the total number of studies involved in these NMAs was relatively small, which led to wide 95% CIs. Second, the included studies varied greatly in participants, intervention measures, control measures, and other factors that directly affected the intervention effect, which made data aggregation difficult. Especially in studies involving non-pharmacological interventions, even the same interventions, such as health education, mostly differed in form across studies which led to clinical heterogeneity in results. There was also some statistical heterogeneity in the design of different studies, such as the follow-up time of intervention, data collection, quality control, and other specific research methods. Third, sensitivity analysis could not be carried out further because of the low quality of the included studies, especially the bias in blinding, allocation consideration, and incomplete outcome data, probably because these studies involved non-pharmaceutical interventions such as health education and had difficulties in achieving total blindness. It was worth noting that there were also studies that fail to achieve the most basic blinding of outcome assessments. Thus, we suggest that it is necessary to adopt blind placebo interventions or sham procedures that differ from experimental interventions and ensure the blinding of intervention effectiveness assessments in future vaccination trails, even if total blinding cannot be achieved. Fourth, some outcomes suggested that there was certain publication bias. Moreover, all the studies included in this paper were conducted prior to the COVID-19 pandemic, so we are not able to evaluate the post-pandemic attitudes towards vaccination, which is a focus that deserves further exploration. In addition, even though we found that a comprehensive intervention strategy could achieve better results, there was still a lack of analysis on the cost of implementing interventions. Therefore, the authorities should choose a plan with a reasonable input–output ratio, taking account the actual situation of the local economic development level and health service resources.

Conclusion

The evidence to date suggests that health education could be beneficial to increase influenza/pneumonia vaccination rates among older adults. A combination of health education and other interventions was also associated with relatively higher improvement of vaccination rates. Without considering the economic cost, health education + centralised reminder + onsite vaccination may potentially be a more effective intervention strategy. However, all the evidence level of findings was low or very low due to the substantial heterogeneity and publication bias among studies which requires further substantiation through more high-quality large-scale RCTs.

Abbreviations

CBM: China Biology Medicine; CI: Confidence interval; CNKI: China National Knowledge Infrastructure; CR: Centralized Reminder; GRADE: Grading of Recommendations Assessment, Development and Evaluation; HE: Health Education; HE + CR + OV: Health education + Centralized reminder + Onsite vaccination; HE + OV: Health Education + Onsite Vaccination; HPQ: Health prescriptions based on questionnaire; NI + OV: Nursing interventions + Onsite vaccination; NMA: Network Meta-Analysis; OR: Odds Ratio; PRISMA: Preferred Reporting Items for Systematic Reviews and Meta-analyses; RCT: Randomized controlled trials; RN: Routine notification; ROB: Risk of Bias; SD: Standard deviation.

Declaration of Conflicts of Interest

None.

Declaration of Sources of Funding

This work was supported by the Xinglin Scholars Program of Chengdu University of Traditional Chinese Medicine (XSGG2020006), the Natural Science Foundation of Sichuan Collaborative Innovation Center of Elderly Care and Health (YLZBZ2009), the Medical Science and Technology Research Fund project of Guangdong Province (C2021106), and the Shenzhen Science and Technology Innovation Commission project (JSGG20210901145406020).

Data Availability

No ethical approval or patient consent was required in that all analyses were conducted based on previously published studies.

References

Author notes