A systematic search was executed across the PubMed, Web of Science, and Embase databases to gather all relevant studies examining the relationship between TAVR and permanent pacemaker implantation outcomes. The search keywords were as follows: (“Transcatheter Aortic Valve Replacement” OR “Transcatheter Aortic Valve Implantation” OR “Transcatheter Aortic Valve Insertion” OR “TAVR” OR “TAVI”) AND (“pacemaker implantation” OR “permanent pacemaker” OR “pacemaker”) AND (“postoperative complications” OR “prognosis” OR “outcome” OR “risk factors” OR “predictors”). The search covered the literature published up to January 2024. Additionally, we also manually searched reference lists of included articles and relevant reviews, and scanned preprint servers (medRxiv, ResearchSquare) for unpublished studies meeting eligibility criteria. All retrieved records were managed in Endnote software, with deduplication algorithms applied before screening.

We included original studies that met the following criteria: (1) Population: Patients undergoing TAVR. (2) Outcome: Studies explicitly reporting of PPM implantation rates within 30 days post-procedure. (3) Design: Randomized controlled trials (RCTs) or observational studies (retrospective cohorts, Newcastle-Ottawa Scale (NOS) scores $\ge$7) with multivariable adjustment for confounders. (4) Risk Analysis: Examination of at least one predefined risk factor (e.g., anatomical, procedural, or electrophysiological variables) associated with 30-day PPM implantation.

Studies were excluded if they (1) involved non-human subjects or focused on basic science mechanisms; (2) lacked clear exclusion criteria for patients with preexisting pacemakers—a critical safeguard against selection bias; (3) compared TAVR with surgical valve replacement or evaluated valve brands without analyzing PPM risk factors; or (4) used aggregated public registry data, which may include duplicate individual patient records from primary studies. To minimize heterogeneity, we also excluded non-English publications and studies reporting outcomes beyond 30 days, as late conduction disturbances often reflected distinct pathophysiological mechanisms.

Two investigators independently performed a dual-phase screening process: initial title/abstract review followed by full-text assessment to determine study eligibility. Data extraction was restricted to articles meeting predefined quality thresholds. From eligible studies, we systematically extracted the following variables: (1) Study identifiers (first author, publication year); (2) Design (prospective/retrospective cohort, RCT); (3) Cohort characteristics (sample size, age, sex distribution, STS score); (4) Quantitative outcomes (number of pacemaker implantations within 30 days post-TAVR).

All extractions were conducted independently by two reviewers using standardized electronic forms. Discrepancies in screening decisions or data interpretation were resolved through iterative discussion, with unresolved cases adjudicated by a third senior investigator. Inter-rater agreement was quantified using Cohen’s $\kappa$ coefficient ($\kappa$ = 0.91 for full-text eligibility).

Two independent investigators evaluated methodological quality using standardized criteria. Cohort studies were assessed with the NOS, scoring selection bias (e.g., cohort representativeness), comparability (adjustment for age/comorbidities), and outcome validity (follow-up adequacy) on a 9-point scale; studies scoring 7 were excluded.

Randomized trials were appraised via the Cochrane Risk of Bias Tool (RoB 2.0, version 1, August 2019; Cochrane Methods Group, London, UK), examining randomization integrity, intervention adherence, missing data handling, outcome measurement consistency, and selective reporting-trials with $\ge$3 high-risk domains were excluded. Disagreements (initially 12% of assessments) were resolved through consensus discussions, with unresolved cases finalized by a third investigators.

Categorical variables were summarized as counts and proportions (%), while continuous variables were reported as means with standard deviations (SD) or medians with interquartile ranges (IQR) based on distribution normality. We employed a random-effects model (DerSimonian-Laird estimator) to pool adjusted odds ratios (ORs) and 95% confidence intervals (CIs), prioritizing this approach to account for anticipated clinical and methodological heterogeneity across studies. Between-study heterogeneity was quantified using Cochran’s Q test (significance threshold: p 0.10) and the I² statistic, with I² values interpreted as follows: $\le$25% (low), 25–50% (moderate), $>$50% (high).

To assess small-study effects and publication bias, we generated funnel plots complemented by Egger’s regression test for asymmetry (p 0.05 indicating significance). Prespecified sensitivity analyses excluded studies contributing disproportionately to heterogeneity (influence analysis) and those with NOS scores 8. All analyses were conducted using R software (version 4.3.1, metafor and dmetar packages, R Foundation for Statistical Computing, Vienna, Austria), with two-tailed p 0.05 defining statistical significance.

Using the aforementioned search strategy, a preliminary identification of 8218 articles was performed. After removing duplicates, 4945 unique articles remained. Following an assessment of article type, abstracts, and keywords, 3994 articles were excluded because of irrelevant topics or because they were literature reviews, conference abstracts, editorial letters, or irrelevant studies. Full-text assessment was conducted on the remaining 951 articles. Subsequently, 802 articles were excluded as they did not meet the inclusion criteria. Finally, a quality assessment was performed on the remaining articles. We conducted a meta-analysis on the risk factors for PPM implantation within 30 days of TAVR, including only those factors that were evaluated in three or more studies. Fig. 1 shows the PRISMA flow diagram.

Fig. 1.

PRISMA flow diagram for the study selection process. PRISMA, Preferred Reporting Items for Systematic Reviews and Meta-Analyses.

Ultimately, we included 82 studies (Table 1, Ref. [4, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86]) involving 124,808 patients. Among these, 21,919 (17.5%) required PPM implantation within 30 days of TAVR. A total of 29,443 patients (23.6%) received self-expanding prostheses, while 42,414 (34.0%) received balloon-dilated prostheses. The remaining patients either did not have the specific valve type reported or were treated with other types of valves. The average PPM implantation rate was 20.4% for self-expanding prostheses and 12.8% for balloon-dilated prostheses. The average age of the included population was 82 $\pm$ 3 years, and 50.6% were males. The mean Society of Thoracic Surgeons risk score was 4.6 $\pm$ 4.5, and the average body mass index (BMI) was 26.5 $\pm$ 1.1 kg/m². Additionally, 83.0% and 31.6% of the patients had hypertension and diabetes, respectively, and 52.4%, 23.3%, 28.3%, and 33.6% had a history of coronary artery disease, chronic obstructive pulmonary disease, atrial fibrillation, and chronic kidney disease, respectively.

Several baseline patient factors were significantly associated with PPM implantation. RBBB showed a strong association with higher odds of PPM implantation (OR = 5.48, 95% CI: 4.52–6.64, p 0.0001), with moderate heterogeneity (I² = 37.74). Sensitivity analysis (Fig. 2) confirmed the robustness of this finding as the exclusion of individual studies did not substantially alter the overall OR, which remained between 5.0 and 6.0. Furthermore, we used funnel plots to assess publication bias and found that, although some asymmetry was observed, there was no significant evidence of publication bias for RBBB (Fig. 3). Additional funnel plots and sensitivity analyses for other risk factors are available in the Supplementary Materials. The first-degree AVB was also significantly associated with PPM implantation (OR = 2.30, 95% CI: 1.82–2.90, p 0.0001), with no observed heterogeneity (I² = 0). There was no significant publication bias, and the sensitivity analysis confirmed the robustness of this association. Baseline left bundle branch block (LBBB) showed no significant association (OR = 1.43, 95% CI: 0.67–3.07, p = 0.3599), but had high heterogeneity (I² = 78.15). There was no significant publication bias; however, sensitivity analysis indicated poor robustness. Longer membranous septum (MS) length was significantly associated with PPM implantation (OR = 0.78, 95% CI: 0.66–0.93, p = 0.0065), with moderate heterogeneity (I² = 78.13). There was no significant publication bias, and the sensitivity analysis confirmed the robustness of this association. Mitral annular calcification (MAC) was not significantly associated with PPM implantation (OR = 1.06, 95% CI: 0.75–1.49, p = 0.7560). Potential publication bias was observed, and the sensitivity analysis indicated poor robustness. Male gender was also not significantly associated with PPM implantation (OR = 0.88, 95% CI: 0.65–1.19, p = 0.3963). There was no significant publication bias; however, sensitivity analysis indicated poor robustness.

Fig. 2.

Sensitivity analysis for RBBB. The blue dots represent the ORs obtained after excluding each study, with the vertical lines indicating the 95% confidence intervals. The red dashed line represents the overall OR, showing consistent results across study exclusions. RBBB, Right Bundle Branch Block; OR, odds ratio.

Fig. 3.

Funnel plot for assessing publication bias in the association between RBBB and PPM implantation within 30 days post-TAVR. PPM, permanent pacemaker; TAVR, transcatheter aortic valve replacement.

Several procedural factors have been identified as significant risk factors for PPM implantation. Increased implant depth per mm was significantly associated with higher odds of PPM implantation (OR = 1.20, 95% CI: 1.13–1.28, p 0.0001) as well as low implant depth (OR = 1.18, 95% CI: 1.12–1.24, p 0.0001), both showing low to no heterogeneity. There was no significant publication bias, and the sensitivity analysis confirmed the robustness of these associations. The use of self-expanding valve was significantly associated with increased odds of PPM implantation (OR = 2.59, 95% CI: 2.06–3.27, p 0.0001), with substantial heterogeneity (I² = 60.13). Potential publication bias was observed, but sensitivity analysis confirmed the robustness of this association.

Additionally, the use of the COT significantly reduced the risk of PPM implantation (OR = 0.45, 95% CI: 0.35–0.58, p 0.0001), with no observed heterogeneity (I² = 0). There was no significant publication bias, and the sensitivity analysis confirmed the robustness of this finding. The difference between MS length and implantation depth ($\mathrm{\Delta }$MSID) was also significantly associated with PPM implantation (OR, 1.36; 95% CI, 1.22–1.50, p 0.0001), with high heterogeneity (I² = 41.82). There was no significant publication bias, and the sensitivity analysis confirmed the robustness of this association. Predilation was another significant risk factor (OR = 1.37, 95% CI: 1.10–1.71, p = 0.0045), with no observed heterogeneity. Potential publication bias was noted, and the sensitivity analysis indicated poor robustness.

Only one post-procedural risk factor met the inclusion criteria. New-onset LBBB was identified as a significant post-procedural risk factor for PPM implantation (OR = 2.26, 95% CI: 1.66–3.07, p 0.0001), with no observed heterogeneity (I² = 0). This finding highlights the importance of monitoring for new-onset LBBB after TAVR. However, it should be noted that a potential publication bias was observed, and the sensitivity analysis indicated a limited robustness of this association.

A temporal trend analysis based on the year of publication was performed to evaluate changes in the 30-day PPM implantation rate following TAVR (Fig. 4). Meta-regression revealed a significant decreasing trend in overall PPM implantation rates over time (coefficient = –0.008, 95% CI: –0.015 to –0.001, p = 0.023). Specifically, a notable decline was observed in the self-expanding valve subgroup (coefficient = –1.571, 95% CI: –2.620 to –0.521, p = 0.004), whereas no significant temporal change was observed for the balloon-expandable valve subgroup (coefficient = –0.248, 95% CI: –1.408 to 0.911, p = 0.666).

Fig. 4.

Temporal trend in 30-Day PPM implantation rates after TAVR.

Subgroup analysis by study design was performed to investigate the potential impact of different study methodologies (RCT, prospective cohort, retrospective cohort) on reported risk factors for PPM implantation. The overall PPM rates did not significantly differ among these study types (Kruskal-Wallis test: statistic = 4.00, p = 0.261). Furthermore, subgroup analyses for specific risk factors (e.g., baseline RBBB, self-expanding valve use, increased implantation depth, baseline LBBB, low implant depth, MAC, and pre-dilation) did not reveal significant heterogeneity between RCTs and observational studies, with all risk factors demonstrating non-significant differences across different study designs (all p $>$ 0.05).

This study aimed to investigate the risk factors for PPM implantation within 30 days of TAVR to provide evidence for perioperative management [87]. Our findings indicated that baseline factors such as RBBB and first-degree AVB were associated with an increased risk of PPM implantation, whereas a longer MS length was associated with a lower risk. Procedural factors, including increased implant depth, low implant depth, use of self-expanding valves, and predilation, were associated with a higher risk of PPM implantation. Conversely, a lower $\mathrm{\Delta }$MSID and the use of the COT could reduce the risk of PPM implantation. Postoperative new-onset LBBB was a risk factor for increased PPM implantation. Sex, MAC, and baseline LBBB were not associated with PPM implantation. Fig. 5 illustrates the overall forest plot of various risk factors associated with PPM implantation within 30 days post-TAVR.

Fig. 5.

Overall forest plot for various risk factors associated with PPM implantation within 30 days post-TAVR using random effects model. SEV, Self-Expanding Valve; AVB, Atrioventricular Block; MAC, Mitral Annular Calcification; COT, Cusp Overlap Technique; LBBB, Left Bundle Branch Block; $\mathrm{\Delta }$ MSID, Difference Between Membranous Septum Length and Implantation Depth; MS, Membranous Septum; New onset LBBB, New Onset Left Bundle Branch Block.

Identifying high-risk patients for PPM implantation before intervention is crucial, as it can significantly influence the treatment strategy and patient’s prognosis. Our meta-analysis found that baseline RBBB increased the risk of PPM implantation by approximately 5.5 times, which is much higher than that of other risk factors. This increased risk was likely because valve implantation could easily damage the left bundle branch. If patients with a preexisting RBBB developed LBBB after TAVR, the incidence of chronic arrhythmias and high-grade AVB increased [6, 7]. However, our study found that the baseline LBBB was not associated with an increased need for PPM implantation after TAVR. This finding may further indicated why patients with a baseline RBBB were more prone to PPM implantation due to the damage to the LBBB caused by the valve deployment and positioning during the procedure. Additionally, first-degree AVB significantly increased the risk of PPM implantation, indicating that preoperative attention to conduction abnormalities was necessary.

Patients with longer MS had a lower risk of PPM implantation, which might be related to anatomical factors. Typically, the His bundle originates from the atrioventricular node, traverses the central fibrous body, and extends into the membranous septum, coursing below the junction of the noncoronary and right coronary cusps, with a total length of approximately 20 mm [88]. As the His bundle and left bundle branch are close to the aortic annulus, some conduction abnormalities during surgery are secondary to mechanical damage to the aortic root, leading to tissue inflammation, edema, or ischemia [89]. A longer MS suggests a greater distance from the annulus to the His bundle, reducing the likelihood of valve-induced compression and thus lowering the risk of conduction block [8, 9].

Interestingly, other meta-analyses have reported that male patients have a higher risk of PPM implantation after TAVR [10, 90], whereas some have found that female patients are more likely to develop new-onset LBBB [6]. However, in our study, which analyzed studies that adjusted for sex in multivariate analyses, we found that sex was not associated with the risk for PPM implantation. These results were consistent across the included studies, indicating no heterogeneity. Some studies have suggested that the higher risk in males may be due to the more frequent use of oversized valves and the higher prevalence of baseline comorbidities [91]. The effect of sex on the risk of PPM implantation remains controversial and warrants further investigation.

Except for baseline factors, procedure-related risk factors were crucial for outcomes. Our findings were consistent with those of previous studies indicating that the use of self-expanding valves significantly increased the risk of PPM implantation after TAVR [6, 89, 90]. This increase might be due to mechanical damage or pressure exerted on the conduction system, leading to tissue inflammation, ischemia, edema, and subsequent conduction abnormalities [2, 11, 12, 92]. Similarly, predilation procedures also elevated the risk of conduction abnormalities by 1.37 times, although the funnel plot indicated potential publication bias for predilation. Predilation should not be performed routinely unless necessary. Valve implantation depth was another critical factor that influenced outcomes. As previously mentioned, the cardiac conduction system was associated with the MS length. Our meta-analysis indicated that valve implantation depth was closely related to the occurrence of conduction abnormalities, which supported the mechanism underlying these post-TAVR complications. Consequently, the measurement of $\mathrm{\Delta }$MSID has been proposed to predict the risk of PPM implantation. Both MS length and valve implantation depth could be measured, and our meta-analysis confirmed that $\mathrm{\Delta }$MSID was significantly associated with PPM implantation, which provided valuable insights into treatment strategies. Preoperative CT assessment of MS length allows calculation of $\mathrm{\Delta }$MSID, which guides individualized valve depth positioning.

In recent years, the COT has been proposed as a novel projection method for valve deployment. This technique involved overlapping the right and left coronary cusps to eliminate parallax, thereby facilitating accurate assessment of implantation depth and reducing the risk of conduction abnormalities [93]. Our study demonstrated that the COT significantly reduced the risk of PPM implantation after TAVR, with no heterogeneity observed among the included studies. COT is beneficial for high-risk patients, especially those with RBBB or first-degree AVB, to mitigate conduction injury.

Among the postoperative factors, only new-onset LBBB met the inclusion criteria. New-onset LBBB was common after TAVR, and we found that it increased the risk of PPM implantation by 2.26 times. Although one-third of patients could reach a resolution of LBBB within 30 days postoperatively as myocardial injury, inflammation, or edema subsided, it is important to note that new-onset LBBB implied the cardiac conduction system had been affected [94]. Therefore, monitoring for conduction block complications in these patients should be warranted, and continuous electrocardiogram monitoring might be reasonable for this patient group [87, 95]. However, it should be paid attention that potential publication bias was observed in our study and the robustness of the findings was limited, necessitating cautious interpretation and further research on the relationship between new-onset LBBB and the need for PPM implantation after TAVR.

The temporal trend analysis demonstrates a significant overall decline in the incidence of PPM implantation following TAVR over recent years, particularly pronounced in self-expanding valve cohorts. This trend likely reflects technological advancements, such as the adoption of the cusp overlap technique, improved procedural optimization, and iterative upgrades in valve design, all contributing to more precise valve positioning and reduced conduction disturbances [93]. However, no significant temporal change was identified in balloon-expandable valve cohorts, possibly indicating a plateau in technological advancements or a consistent patient selection approach in this subgroup.

Our study has several limitations. First, as this is the first meta-analysis to focus on the perioperative period of TAVR, we excluded high-quality studies whose endpoints did not align with our criteria. Second, we only included studies that performed multivariate adjustments for risk factors to eliminate confounding effects. Consequently, compared with previous meta-analyses, only 13 factors met our criteria; however, this provided stronger evidence regarding the impact of these factors on PPM implantation after TAVR. Third, although some variables such as BMI [13, 14], choice of oversized valves [4], and changes in QRS duration [15, 16], have been identified in previous studies as potentially related to conduction abnormalities, we required a minimum of three studies for each variable to ensure reasonable evaluation. Owing to differences in definitions among studies, some variables were ultimately not included in our analysis [96]. Lastly, although the number of eligible RCTs included in our meta-analysis was limited, we employed rigorous quality assessment tools for observational studies and conducted subgroup analyses based on study design. The results showed that the estimated effects of risk factors were largely consistent across study types, thereby strengthening the reliability of our overall findings.