Factors influencing intraoperative and postoperative complication occurrence: A series of 1135 periodontal and implant‐related surgeries

Abstract Objectives In periodontology, it is widely recognized that evidence characterizing the incidence and effect of treatment complications is lacking. The objective of this study was to assess the influence of operator‐, procedure‐, patient‐, and site‐associated factors on intraoperative and postoperative complication occurrence. Material and Methods A single investigator reviewed records of patients treated by eight periodontics residents from July 2018 through June 2022. For each procedure, the investigator recorded each intraoperative and postoperative complication or indicated that no complication had occurred. These outcomes were analyzed against a panel of explanatory covariates. In addition, the severity of each postoperative complication was assessed using a standardized grading system. Results A total of 1135 procedures were included in the analysis. Intraoperative and postoperative complications were identified in 2.8% and 15.2% of procedures, respectively. The most common intraoperative complications were Schneiderian membrane perforation (1.3%) and gingival flap perforation/tear (1%), and the most common postoperative complications were dentin hypersensitivity (2.6%), excessive pain (2.5%), and infection (2.2%). Subepithelial connective tissue graft (odds ratio [OR]: 3.2, 95% confidence interval [CI]: 1.6, 6.1; p < .001), guided bone regeneration (OR: 3.0, 95% CI: 1.4, 6.5; p = .004), and guided bone regeneration with implant placement (OR: 3.1, 95% CI: 1.3, 7.6; p = .011) were associated with higher odds of postoperative complication, whereas lateral sinus elevation (OR: 102.5, 95% CI: 12.3, 852.9; p < .001), transalveolar sinus elevation (OR: 22.4, 95% CI: 2.2, 224.5; p = .008), open flap debridement (OR: 36.4, 95% CI: 3.0, 440.7; p = .005), and surgically facilitated orthodontic therapy (OR: 20.5, 95% CI: 1.2, 358.4; p = .039) were associated with higher odds of intraoperative complication occurrence. Conclusions Consistent with previous reports, procedure type appears to be the predominant factor driving complication occurrence. As analyses of treatment complications increase, individualized risk‐benefit assessments will become progressively meaningful for patients.


| MATERIALS AND METHODS
The Dwight David Eisenhower Army Medical Center Human Research Protections Office reviewed and approved this protocol (#21-14188/945052) in accordance with the exemption criteria described in 32CFR §219.104(d),Category 4. Data were collected in such a manner that the identities of human subjects were not readily discernible directly or through identifiers linked with subjects.The

| Exclusion criteria
Patients who received a surgical procedure other than the defined procedure types, patients with ambiguous or incomplete records, and patients with less than 6 weeks follow-up were excluded from this analysis.
The primary outcomes of interest consisted of intraoperative and postoperative complication occurrence.These outcome variables were compared against a panel of operator-, procedure-, patient-, and site-related covariates.Specifically, independent variables included operator, year of training, patient age, patient gender, diabetic status, current smoking status, region (anterior or posterior), site type (single tooth position or multiple adjacent tooth positions), peri-operative antibiotic use, and procedure type (defined in the inclusion criteria).In addition to recording occurrence or absence of a postoperative complication, the investigator scored postoperative complications according to the grading scale of Askar and colleagues (2019).Smoking status and diabetic status each had two levelscurrent smoker versus nonsmoker and diabetic versus nondiabetic.A diagnosis of diabetes was confirmed in the patient record.Patients who smoked at least one cigarette per day were considered smokers.Postoperative pain was considered excessive if the provider indicated in the patient record that the pain exceeded the expected level, extended the patient's convalescence period due to discomfort, or prescribed additional analgesics.Postoperative swelling was considered a complication if the provider indicated in the patient record that swelling exceeded the expected level, persisted for at least 2 weeks, or required additional appointments for monitoring in the early postoperative period.Postoperative bleeding was considered a complication if the patient returned to the clinic with bleeding requiring an intervention to achieve hemostasis or if the patient reported bleeding to the surgeon telephonically but was able to achieve hemostasis with gauze and pressure without returning to the clinic.Implant failure within 6 months of placement was considered a postoperative complication for the purpose of this analysis.

| Statistical analyses
Descriptive and inferential statistics were calculated for all variables using IBM SPSS Statistics for Windows v.28, and statistical significance was assessed at an alpha level of 0.05.Binomial logistic regression analyses were used to ascertain the effects of operator-, procedure-, patient-, and site-related explanatory variables (listed above) on intraoperative and postoperative complication occurrence.For each regression model, omnibus tests were used to assess model fit.Linearity of the continuous variable (age) with respect to the logit of the dependent variables was assessed via the Box-Tidwell procedure.Odds ratios (ORs) and 95% confidence intervals (CIs) were determined.Use of binomial logistic regression required definition of reference categories for each categorical independent variable-procedure type, antibiotic, and operator.
Delayed implant placement, a relatively simple and low-risk procedure, was selected as the reference category for procedure type."No antibiotic" served as a non-intervention reference category for antibiotics, and Operator 1 served as the reference category for the operator variable.

| RESULTS
Table 1 summarizes the study sample and presents descriptive statistics for the dependent and independent variables, and Table 2 presents the frequencies of intraoperative and postoperative complications for each procedure type.In this study, male patients outnumbered female patients by a ratio of nearly 4:1.Less than 10% of the patients were smokers, and only 3.5% of patients were diabetic.In 680 (59.9%) procedures, the patient received no perioperative antibiotics.Second-year and third-year residents performed 507 (44.7%) and 628 (55.3%) of the procedures, respectively.
Of 1135 evaluated procedures, 31 (2.8%) resulted in an intraoperative complication and 173 (15.2%) resulted in a postoperative complication.Of the 173 postoperative complications, 55 (31.8%) were localized complications accompanied by no adverse effects on the success of the procedure (grade 1), 79 (45.7%) were localized complications that did adversely influence the success of the procedure (grade 2), and 39 (22.5%) were localized or systemic complications that temporarily impaired the patient's daily routine (grade 3).No complication in this sample required hospitalization of the patient (grade 4) or resulted in irreversible damage to an anatomic structure (grade 5).
A binomial logistic regression analysis was completed to determine whether operator-, patient

| DISCUSSION
This investigation was conducted to evaluate the effect of operator-, procedure-, patient-, and site-related factors on complication occurrence following periodontal and implant surgeries.Increased reporting and analysis of treatment complications to shape clinical decisionmaking within periodontics has been encouraged (Chambrone & Zucchelli, 2022;Park et al., 2011;Zucchelli et al., 2023).The most impactful data on complications and adverse outcomes will originate from results of controlled clinical research.Observational investigations, including the current study, are subject to biases and confounding factors that may lead to spurious associations.Nevertheless, this study involved a relatively large number of observations, and a retrospective analysis of complication occurrence was accomplished without risk to any patient.Moreover, findings from the present study can be placed in context of previous studies reporting complication occurrence in other patient samples (Askar et al., 2019;Barone et al., 2006;Fontana et al., 2011;Hernández-Alfaro et al., 2008;Lim et al., 2018;Moreno Vazquez et al., 2014;Powell et al., 2005;Sakkas et al., 2016;Schwartz-Arad et al., 2004;Zucchelli et al., 2010).

| Postoperative complication occurrence
Although "operator" was ultimately not found to be a statistically significant predictor of postoperative complication occurrence in this | 7 of 14 study sample, this variable was included in the final regression model due to statitistical significance during hierarchical omnibus testing.
Perceived clinical skill has been reported as the primary determinant of patient referrals from general practitioners to periodontists (Kraatz et al., 2017;Park et al., 2011;Zemanovich et al., 2006), and in various dental and medical disciplines, operator skill/experience has been shown to significantly influence treatment outcomes across a wide range of procedure types (Carter, 2003;Cushen & Turkyilmaz, 2013;Kozlovsky et al., 2018;Stulberg et al., 2020).In one study involving general surgeons, operator skill was estimated to account for 25% of the variance in treatment outcomes (Stulberg et al., 2020).In uncomfortable (Wessel & Tatakis, 2008).Overthinning the superficial tissue at the donor site can result in necrosis and exposure of underlying connective tissue, permitting topical irritation of the palatal wound (Zucchelli et al., 2010).Periodontists in training may be more likely to overthin the superficial tissue at the palatal harvest site.In addition, palatal graft dimensions have been shown to correlate with postoperative pain (Zucchelli et al., 2010).It is possible that SCTG dimensions influenced the observed postoperative complication occurrence in this study.
GBR and GBR with implant placement were treated as separate procedures for the purpose of this analysis.GBR procedures acomplished in advance of implant surgery may tend to involve larger alveolar ridge deficiencies, whereas small ridge deficiencies may more likely permit simultaneous implant installation.Nevertheless, both GBR alone and GBR with implant placement were associated with increased risk of postoperative complication occurrence in this study.In this sample, the most common complications in GBR procedures with or without implant placement were excessive pain, postoperative infection, and membrane exposure.Multiple authors have associated these complications with GBR (Elgali et al., 2017;Fontana et al., 2011;Lim et al., 2018).Reported soft tissue complication rates following GBR vary substantially-ranging from 0% to 45% (Lim et al., 2018).Membrane type (absorbable vs. nonabsorbable) (Elgali et al., 2017;Urban, 2017) and operator skill (Lim et al., 2018) have been suggested as factors influencing this metric.In particular, chemically cross-linked collagen membranes and nonabsorbable polytetrafluoroethylene membranes have been associated with high exposure risk (Elgali et al., 2017).
The postoperative complication findings observed in the present study are comparable with results of a previous retrospective investigation involving 3900 procedures (Askar et al., 2019).The reported overall postoperative complication rates were 20.1% and 15.2% in the previous and present studies, respectively (Askar et al., 2019).The surgical procedures evaluated in the two cohorts were similar, although the present analysis included SFOT, ADM, ARP, and LPT rather than third molar surgery.Askar and colleagues (2019) included postoperative chlorhexidine rinse as a variable (Askar et al., 2019).However, in the present study, 0.12% chlorhexidine gluconate was utilized for plaque control following every surgical intervention, until the patient's normal oral hygiene regimen could be reinstated.The present study also categorized compound procedures such as implant placement + GBR, whereas the previous study excluded patients receiving more than one surgical intervention.
Additionally, the present study analyzed consecutive surgical procedures accomplished by eight periodontics residents.Thus, the number of procedures assessed in each surgical category varied.In contrast, the previous study involved assessment of 300 records for each procedure of interest, establishing groups of equal size before completing the analysis.In the previous study, the procedures were predominantly performed by supervised residents/students in various dental disciplines, while the procedures in the present study were performed by supervised periodontics residents only.Whereas Askar and colleagues analyzed surgical procedures performed in a university setting, patients in the present study were predominantly active duty military service members, with a small number of military retirees.Despite dissimilar patient populations and substantive methological differences, the two most common postoperative complications in both cohorts were dentin hypersensitivity and excessive pain.
In the present study, smoking did not predict complication occurrence.Smokers with periodontitis are known to respond less favorably to surgical and nonsurgical periodontal therapy (Preber & Bergström, 1986;Tonetti et al., 1995).Cigarette smoking has been associated with reduced implant survival (Anner et al., 2010;Heitz-Mayfield & Huynh-Ba, 2009) complications following implant surgery (Schwartz-Arad et al., 2002), and inferior clinical outcomes in multiple hard and soft tissue grafting procedures (Erley et al., 2006;Horváth et al., 2013;Lindfors et al., 2010;Miller, 1987;Patel et al., 2012;Zitzmann et al., 1999).Increased risk of alveolar osteitis and excessive pain has also been found in cigarette smokers receiving third molar surgery (Schwartz-Arad et al., 2017).Procedures involving implanted autogenous tissues and biomaterials appear particularly Postoperative complication percentage by operator.In a binomial logistic regression model including operator as the only variable, Operator 6 was associated with significantly lower odds of postoperative complication occurrence.Thus, operator was included as an independent variable in the final model.However, in the final model, this effect was not statistically significant.The number in each bar indicates the actual number of events observed.
susceptible to the detrimental effects of smoking.For example, the variance in root coverage outcomes for smokers versus nonsmokers appears pronounced in SCTG-based procedures but may be less significant when the coronally advanced flap technique is employed without implanted autogenous or allogeneic tissue (Chambrone et al., 2009;Chambrone & Tatakis, 2015;Erley et al., 2006)
OFD was associated with increased risk of an intraoperative complication in the present study.This procedure is typically planned when periodontal defect severity and configuration are unfavorable for regenerative or resective surgery and when patient-related factors dictate a simplified surgical approach.In residency, less complex procedures such as OFD may also tend to be performed by less experienced clinicians.In the present study, the observed association between OFD and intraoperative complication occurrence may relate to the low number of observations.Two complications were recorded in 11 OFD procedures.One procedure involved a tear in the mucoperiosteal flap.The other procedure was aborted and rescheduled due to anxiety-related tachycardia.

| CONCLUSIONS
In this study, SES and OFD were associated with intraoperative complication occurrence, whereas SCTG and GBR were associated procedures evaluated in this study were performed by eight residents in the Department of Periodontics, Army Postgraduate Dental School, Postgraduate Dental College, Fort Eisenhower, Georgia, between July 1, 2018, and June 10, 2022.Before treatment, each patient completed an informed consent process involving verbal and written components.This investigation complies with the Declaration of Helsinki and the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines.
Intraoperative complication types defined in this study included mucoperiosteal flap perforation/tear, Schneiderian membrane perforation, retained root fragment, excessive intraoperative bleeding, thermal injury, iatrogenic damage to adjacent structures, and the need to abort surgery for other reasons.Postoperative complication types defined in this study included dentin hypersensitivity, excessive postoperative swelling, postoperative bleeding, wound dehiscence, membrane exposure, postoperative infection, graft necrosis, excessive postoperative pain, ecchymosis, altered neurosensory function, sinusitis, trismus, urticaria, gastrointestinal disturbance, and early implant failure (osseointegration failure).

(
66.7%) Schneiderian membrane perforations occurred during lateral SES, three (20.0%)during transalveolar SES, and two (13.3%)during tooth extraction with ARP.Five (45.4%) mucoperiosteal flap perforations/tears occurred during SCTG procedures, two (18.2%)during implant placement with simultaneous GBR, and two (18.2%)during transalveolar SES.A binomial logistic regression analysis was completed to determine whether operator-, patient-, site-, and procedure-related factors predicted postoperative complication occurrence.The model was built in a hierarchical fashion.Omnibus tests of model coefficients were used to eliminate from the model factors that were not significant predictors of the outcome.Only operator (Figure 2) and procedure type remained in the final model, and in the model including both of these variables, only procedure type was a statistically significant predictor of postoperative complication occurrence (Table contrast, neither year of training nor clinican significantly influenced postoperative complication occurrence in the present study.All evaluated procedures were completed by second-and third-year periodontics residents.Thus, this study did not involve comparisions across a broad spectrum of operator experience.Additionally, each F I G U R E 1 Intraoperative and postoperative complication percentages by procedure type.In binomial logistic regression analyses, lateral SES, transalveolar SES, and open flap debridement were associated with significantly higher odds of intraoperative complication occurrence.Subepithelial connective tissue graft, guided bone regeneration, and implant plus guided bone regeneration were associated with significantly higher odds of postoperative complication occurrence.The number in each bar indicates the actual number of events observed.ADM, acellular dermal matrix; ARP, alveolar ridge preservation; BRG, bone replacement graft; EPG, epithelialized palatal graft; GBR, guided bone regeneration; GTR, guided tissue regeneration; GV, gingivectomy, LPT, laser periodontal therapy; OFD, open flap debridement; SCTG, subepithelial connective tissue graft; SES, sinus elevation surgery; SFOT, surgically facilitated orthodontic therapy.evaluated produre was directly supervised by a board certified periodontist, and all supervising faculty members had similar training and professional experience.These biases may have mitigated the impact of operator skill/experience in this study.Procedure type was the only independent variable identified as a statisically significant predictor of postoperative complication occurrence in this study sample.SCTG was associated with threefold increased odds of postoperative complication, and in this study, the most common complication following SCTG was excessive postoperative pain.Pain from the palatal donor site is a known disadvantage of autogenous soft tissue grafting; however, compared with EPG donor sites, SCTG donor sites are typically less with postoperative complication occurrence.Although the operator appeared to influence postoperative complication occurrence, this variable was not found to be statistically significant in the final logistic regression model.Characteristics of the study sample, the study methodology, and biases unique to this investigation may account for smoking, diabetes, operator experience, and perioperative antibiotic use having no detectable effect.The findings of this investigation indicate that procedure type is the predominant factor driving complication occurrence.Evaluation of treatment complications in various contexts represents an essential input into the clinical decision-making process.As additional treatment complication analyses become available, risk-informed practitioners will increasingly gain the ability to engage patients in meaningful informed consent processes.
T A B L E 2 Binomial logistic regression predicting postoperative complication occurrence.