Association between endodontic, patient‐related factors and severe odontogenic infections; a South Australian retrospective audit

Abstract

Aim

Endodontic conditions are associated with severe odontogenic infections. However, no studies have explicitly explored their role in the infectious process. This study aimed to assess associations between tooth‐related and patient‐related factors resulting in severe odontogenic infections and clinical care needs.

Methodology

A retrospective audit was conducted of all records of patients who presented with severe odontogenic infections requiring admission between 2018 and 2023, at the largest public hospital in South Australia. Pericoronitis, postoperative, and non‐odontogenic infections were excluded. Associations between tooth‐related factors (type, location, presence/absence of caries, restorations, root canal filling, apical periodontitis, and space of infection), patients‐related factors (diabetes and smoking statuses), clinical management (length of admission, intensive care unit admission number of days, and need for general anaesthesia) were explored. χ ², Mann–Whitney's test and Fisher's exact test were used for statistical analysis, and the level of statistical significance was set at 5%.

Results

Of 382 records identified, 189 met the inclusion criteria. Mandibular (95.8%) molars (92.1%) were the most commonly offending teeth. Eight (4.2%) were root canal filled, 27 (14.3%) restored, 166 carious (87.8%), and 181 (95.8%) were associated with an apical radiolucency. Over half the sample were smokers (56.9%) and 16.4% were diabetics. The most common infection involved the submandibular space (56.1%). Intensive care unit admission was required for 157 patients (83.1%), and stay depended on the space of infection, with buccal space having the shortest duration (median 1 day) compared with submandibular (median 4 days) (p < .001). Need for general anaesthesia was associated with Ludwig's angina (n = 3, 100%) (p = .04) and submandibular space infection (n = 104, 98.1%) (p < .001), plus intensive care unit admission (p < .001). Conversely, for buccal space infection, the need for general anaesthesia was lower (n = 5, 23.8%). No further significant differences were found.

Conclusion

Untreated mandibular molars with apical periodontitis were the most frequent tooth‐related findings. Results from the present study suggest associations between the most severe clinical presentations (Ludwig's angina and submandibular infection) and complexity in clinical care. Buccal space infections required the least complex management.

INTRODUCTION

Odontogenic infections are a relatively common occurrence and severe presentations have the potential to cause significant morbidity, and less often, mortality if not treated effectively as recorded in Finland, the United States, and Australia (Seppänen et al., 2008; Shah et al., 2013; Ullah et al., 2023; Uluibau et al., 2005). Amongst the possible presentations, Ludwig's angina is a severe, life‐threatening infection arising if there is bilateral involvement of the submental, sublingual, and submandibular spaces (American Association of Endodontists, 2020). It is characterized by painful swelling of the floor of the mouth, tongue elevation, dysphagia, dysphonia, and (at times) airway compromise. Notably, a significant increase in the number of hospital admissions for surgical treatment of dental infections has been reported in Australia, Germany, and England in the last decades (Fu et al., 2020; Meisgeier et al., 2024; Robertson & Smith 2021).

Oral diseases include dental caries and apical periodontitis. Dental caries can cause pulp necrosis and contamination of the root canal space, with bacteria playing a role in developing both diseases (Manji et al., 2018; Nair, 2004). Dental caries is the most common disease globally, affecting around 35% of the population with permanent dentition (Kassebaum et al., 2015). Similarly, apical periodontitis is an inflammatory condition with a calculated prevalence of 52% worldwide (Tibúrcio‐Machado et al., 2021).

Patient‐related factors may influence odontogenic infections, with immune response disorders increasing the risk of infection dissemination (Cintra et al., 2021). Individuals with one or more systemic conditions have an increased prevalence of apical periodontitis (63%) (Tibúrcio‐Machado et al., 2021). Amongst the various systemic health conditions, diabetes is a major public health issue worldwide affecting more than 422 million people with an increasing occurrence (World Health Organization, 2024). Diabetes is the most common condition studied in relation to apical periodontitis, with a purported bidirectional relationship to prevalence, severity, disease control, and response to treatment (Cintra et al., 2021). The pooled prevalence of individuals with at least one tooth with apical periodontitis in type 2 diabetic patients is 75% (Tibúrcio‐Machado et al., 2021). Poorly controlled diabetes is associated with more severe endodontic infections (Segura‐Egea et al., 2019). In addition, smoking is a risk factor for systemic and oral health that affects the immune response to infection by limiting the host response (Cintra et al., 2021; Segura‐Egea et al., 2015). Smoking causes lung diseases (Larsson & Burgess 2022) and therefore is a crucial comorbidity for acute odontogenic infections, as the latter are life‐threatening by impairing respiration. Notably, smoking is a significant global issue, with an estimated 1.3 billion tobacco users worldwide (World Health Organization, 2023).

The clinical management of acute odontogenic infection includes variables such as admission length, intensive care unit admission, and the need for general anaesthesia (Han et al., 2020). These will vary widely according to the setting, as calculated in the United States and Finland (Ahmad et al., 2013; Gams et al., 2017; Seppänen et al., 2010). However, they are significant, especially if compared with the routine management of tooth‐related comorbidities when delivered in the primary care setting which vary according to the country (Christensen et al., 2013a; Han et al., 2020). High‐risk infections with airway compromise are associated with an increased cost, length of admission, and duration in the intensive care unit (Han et al., 2020). Thus, apart from morbidity, potential mortality, and productivity loss for the patient, these presentations burden the health system.

There has been no literature systematically evaluating tooth‐related factors (i.e. restorations, caries, previous root canal treatment, and apical periodontitis), and assessing their possible association with severe odontogenic infections. Therefore, the aim of this study was to investigate possible associations between tooth, patient‐related, and clinical care needs in patients presenting with severe odontogenic infections in a general hospital in Australia. The null hypothesis was that there are no associations between tooth, patient‐related factors, and the clinical management of severe odontogenic infections requiring hospital admission.

MATERIALS AND METHODS

The present article was prepared in accordance with the PROBE 2023 guidelines for reporting observational studies in endodontics (Nagendrababu et al., 2023).

The study was approved by the University of Adelaide Human (reference number: H‐2024‐078) and the Central Adelaide Local Health Network (reference number: 17725) Research Committees. Due to the retrospective nature of the study, the requirement of obtaining informed consent was waived and was categorized as Quality Assurance.

A retrospective audit on a convenience sample of 382, obtained from screening 2407 consecutive oral and maxillofacial records was performed to identify odontogenic infections admitted to the Royal Adelaide Hospital in Adelaide, South Australia, between January 2018 and December 2023 through admission database records, including relevant imaging. This hospital is the largest public hospital in the region being the ‘super‐site’ for major emergencies in South Australia, such as burns, heart attacks and strokes. Furthermore, it is the only hospital in the state to receive tertiary referrals for maxillo‐facial emergencies and offers no conservative dental treatment whatsoever.

Inclusion criteria included a diagnosis of severe odontogenic infection of endodontic origin and that database records included relevant imaging. Exclusion criteria included non‐infectious diseases (e.g. trauma, cancer, elective surgeries), pathoses of non‐endodontic origin, pericoronitis, and postoperative infections. Individuals self‐identifying as Aboriginal or Torres Strait islanders were excluded.

Data collection was completed by one investigator (AC), a consultant in oral and maxillofacial surgery holding dental and medical degrees. It included patient‐related factors: sex, age, diabetes (yes/no), and smoking status (current/former/no); tooth‐related factors: tooth group (incisors, canine, premolar, or molar) (when multiple teeth were associated with a periapical radiolucency all were counted), and jaw (maxilla/mandible). For the tooth associated with most radiographic findings in each patient, caries (yes/no), restoration (yes/no), root canal treatment (yes/no), periapical radiolucency (yes/no) (based on imaging), and space of infection were collected; finally, clinical management included needs for general anaesthesia (versus local anaesthesia), length of admission and intensive care unit use. A customized and piloted spreadsheet was used.

The data collection was repeated to ensure accuracy, and the intra‐reviewer reliability (percentage of agreement) of the full data collection was calculated.

Chi‐square was used to assess the association between categorical variables. Fisher's exact test was used when a high number of low expected frequencies was observed, and Mann–Whitney's test was used to compare the distribution of an ordinal variable amongst two independent groups, and the level of statistical significance was set at 5%. Statistical analyses were performed using the SPSS 21.0 statistical package (SPSS Inc., Chicago, IL, USA).

RESULTS

From the 382 patients identified with severe odontogenic infections requiring hospital admission, 189 fulfilled the inclusion criteria following the assessment of radiographic admission and operation records, as illustrated in Figure 1. Frequency distribution for patient and tooth‐related variables is presented in Table 1. Figures 2, 3, 4, 5 illustrate common radiographic findings.

FIGURE 1.

Case selection flowchart listing the main reason for exclusion.

TABLE 1.

Frequency distribution for patient, tooth‐related, and clinical management variables.

	Frequency (n)	Percentage (%)
Patient‐related factors
Sex
Male	119	63
Female	70	37
Age (years)
20–29	39	20.6
30–39	44	23.3
40–49	40	21.2
50–59	32	16.9
60–69	20	10.6
70–79	11	5.8
80–89	2	1.1
90–99	1	0.5
Diabetes diagnosis
Yes	31	16.4
No	351	83.6
Smoking status
Non‐smoker	42	32.3
Current smoker	74	56.9
Former smoker	14	10.8
Tooth‐related factors
Tooth type a
Incisor	2	1.1
Canine	6	3.2
Premolar	29	15.3
Molar	174	92.1
Jaw
Maxilla	8	4.2
Mandible	181	95.8
Tooth condition
Caries
Yes	166	87.8
No	23	12.2
Restoration
Yes	27	14.3
No	162	85.7
Root canal treatment
Yes	8	4.2
No	181	95.8
Periapical radiolucency
Yes	181	95.8
No	8	4.2
Space infection b
Submandibular	106	56.1
Buccal	21	11.1
Pterygomandibular	16	8.5
Submasseteric	11	5.8
Parapharyngeal	10	5.3
Submental	9	4.8
Sublingual	6	3.2
Ludwig's angina/multispace	3	1.6
Clinical management
General anaesthesia
Yes	168	88.4
No	21	11.6
Intensive care admission
Yes	157	83.1
No	32	16.9

^a Multiple teeth were associated with one apical pathosis in some records.

^b If frequency per space was ≤2 the data were not reported.

FIGURE 2.

Panoramic radiograph illustrating mandibular left second molar tooth associated with carious lesion and apical pathosis (arrow).

FIGURE 3.

Panoramic radiograph illustrating mandibular left first molar tooth associated with intracoronal restoration and apical pathosis (arrow).

FIGURE 4.

Computed tomography scans illustrating (a) cracked mandibular left second molar tooth (arrow), and (b) periapical radiolucency (arrow).

FIGURE 5.

Panoramic radiograph illustrating mandibular right first molar tooth with coronal restoration and root canal fillings associated with apical pathosis (arrow).

Patients were 63% male and 37% female; 65.1% were younger than 50 years of age. Overall, the diabetes rate was 16.4%, current smokers were 56.9%, with an additional 10.8% being former smokers.

Teeth‐related factors with higher distribution included molars (92.1%), mandibular (95.8%), carious (87.8%), non‐restored (85.7%), not root canal‐treated (95.8%) and periapical radiolucency present (95.8%). Submandibular space infection was a small majority (56.1%), whereas Ludwig's angina/multispace was relatively uncommon (1.6%).

General anaesthesia and intensive care admission frequencies are presented in Table 1. Ludwig's angina had a general anaesthesia rate of 100%, which was 98.1% for submandibular space infection. Median length of admission was 3.9 days (IQR: 2–4), which depended on the space infection; Ludwig's angina and submandibular space infection had increased median stays (11 and 4 days, respectively) (data for other spaces not reported).

Overall, the median intensive care unit admission duration was 1 day (IQR:1–2), with a longer duration for Ludwig's angina (median 4 days). Submental location reduced the length of intensive care unit stay (median 1 day; IQR: 0–1) (data for other spaces not reported).

Associations between factors were as follows: Ludwig's angina with multiple offending teeth including premolars (p = .042), and all subjects were current smokers. Current smokers presented more often with restorations (p = .022) and submandibular space location (p = .060). Caries (p = .010), restorations (p = .002), and root canal‐treated teeth (p = .040) were associated with increasing age.

Need for general anaesthesia was associated with intensive care unit admission (p < .001), presentation of Ludwig's angina (p = .004), and submandibular space infection (p < .001). Intensive care unit admission was associated with buccal or submandibular space infections (p < .001), diabetes (p = .027), and Ludwig's angina (p = .046). Associations were found between length of admission and buccal or submandibular (p < .001), sublingual (p = .016) space infections, and Ludwig's angina (p = .015). Intensive care unit length stays were significantly longer for Ludwig's angina when compared with the other spaces (p = .013); submental location reduced the length of intensive care unit stay (p = .048). No further significant differences were found.

There was excellent agreement in data entry (98.8%) for the data, including qualitative comments.

DISCUSSION

The present study assessed restorative and endodontic statuses of teeth associated with severe odontogenic infections. The vast majority were concurrently carious, unrestored, and not root canal‐treated. The presence of open communication with the oral cavity can explain these findings; however, no direct comparison is feasible as all patients presented with a severe infection of endodontic origin, as per the inclusion criteria of the study. The severity of the presentations included can be highlighted by the intensive care unit admission rate, with more than 80% of patients requiring admission. The most frequent infection was the submandibular space, commonly requiring general anaesthesia. Ludwig's angina was relatively uncommon, although, as expected, was associated with more complex management. Overall, in the presence of several significant differences in the various analyses, the null hypothesis was not rejected.

There is limited evidence regarding long‐term risk of exacerbation of untreated disease with asymptomatic apical periodontitis (Doğramacı and Rossi‐Fedele 2023). Conversely, a longitudinal study assessing exacerbations in adequately root canal‐treated teeth with persistent lesions reported a 5.8% incidence of exacerbations requiring emergency management 20 years post‐treatment, and less severe pain was more common with a 45% incidence after 20 years (Yu et al., 2012). In the present survey, the presence of an adequate root filling and a restoration may be purported as protectors for severe infections due to most teeth not being previously root filled. Notably, valid consent requires discussing what might happen if the proposed treatment is not carried out, including the risk of severe odontogenic infections.

The findings of this audit suggest that most severe odontogenic infections arise as sequelae to untreated caries. Dental caries has been consistently associated with dental infections requiring hospitalization, with rates calculated as 33.8% (Sánchez et al., 2011), 50% (Seppänen et al., 2011), 65% (Flynn et al., 2006), ranging between 59 and 90% (Ullah et al., 2023) or sample described as ‘mostly carious teeth’ (Prabhu & Nirmalkumar, 2019). Regarding the endodontic status of offending teeth in previous studies, the following descriptors and rates have been listed: ‘pulpal origin’ as 81%, of these ‘failed endodontic treatment’ (8%) (Uluibau et al., 2005), ‘post endodontics’ (8%) (Sánchez et al., 2011), ‘nonvital’ (28%) (Moghimi et al., 2013), ‘pulpal origin’ (70%) (Mathew et al., 2012), ‘apical periodontitis’ (87%) being of those presenting with an ‘endodontic treatment’ (18%) (Seppänen et al., 2011). The above findings are expected, considering the high prevalence of caries globally and that caries is the main reason for pulp demise (Kassebaum et al., 2015). Severe odontogenic infections ensue a long disease process often preceded by reoccurring symptoms in the absence of adequate management (Mathew et al., 2012).

Mandibular molars are the most common source of severe odontogenic infection as per the literature (Bottin et al., 2003; Mathew et al., 2012; Moghimi et al., 2013; Opitz et al., 2015; Seppänen et al., 2008; Seppänen et al., 2010; Sánchez et al., 2011). Subsequently, the submandibular space is the most commonly involved (Gams et al., 2017; Seppänen et al., 2008), with the incidence in the present survey being comparable with previous studies as follows: as single space infection 32%–35% (Storoe et al., 2001), 69% (Mathew et al., 2012), 78% (Opitz et al., 2015). Furthermore, multiple space infections often involve the submandibular space (Seppänen et al., 2010; Storoe et al., 2001). A relatively higher percent of incidence in Ludwig's angina has been reported (14.6%) (Mathew et al., 2012), and 52% (Prabhu & Nirmalkumar, 2019), though the latter study included non‐odontogenic aetiologies. The high frequency of mandibular molars as offending teeth leading to infections spreading in the submandibular spaces can be explained by the position of the apices of these offending teeth. The identification of tooth type and other factors facilitates the recognition of risks of severe odontogenic infections and their recurrence.

The role of diabetes has been assessed in the previous studies. One found no significant role despite chronic complications being analysed separately (Allareddy et al., 2010), conversely, two studies have reported longer lengths of admission for diabetics (Gams et al., 2017; Sánchez et al., 2011). In the current study, there were no significant associations between tooth‐related factors and diabetes diagnosis, however, there was a strong tendency to involve untreated caries (p = .052). In previous studies, the percentage of diabetic patients ranged between 3% and 9% (in Australia) (Fu et al., 2020), 8% (Uluibau et al., 2005), 9% (Prabhu & Nirmalkumar, 2019), 11% (Sundararajan et al., 2015), 12% (Gams et al., 2017), 24% (in India) (Mathew et al., 2012) versus 16.4% in the present study. Smoking rates ranging between 58% and 71% (in Australia) (Fu et al., 2020), and 46% (in the United States) (Gams et al., 2017) have been reported, comparable to those of the present study (56.9%). The absence of further significant differences for the above comorbidities may be associated with the use of a convenience sample. Overall, findings highlight the increasing risk of infection requiring hospitalization in individuals with caries and who are current smokers. Considering the high prevalence of diabetes and tobacco smoking worldwide, the results of the present study are significant outside Australia.

The findings related to variables related to the clinical management of acute odontogenic infection in the present survey can be contrasted with those from similar studies, and they highlight the complexity of the management of these patients. It should be underscored that patients normally remain hospitalized until the process is controlled or resolved, there is no airway compromise, and the general health state returns to the pre‐infectional situation (Storoe et al., 2001). The use of general anaesthesia has been reported in India to be 52% (Mathew et al., 2012), whereas in Australia, 74% (Han et al., 2020) or 87% (Uluibau et al., 2005) of patients, thus comparable with the present study (88.4%). Cases requiring general anaesthesia are considered high‐risk (Han et al., 2020).

The length of hospital stay reflects the severity of the condition (Seppänen et al., 2008; Christensen et al., 2013b). The mean length of admission in days reported in comparable studies is as follows: 3.6 (IQR 2.7–5.6) (in Australia) (Sundararajan et al., 2015), 4.18 (Liau et al., 2018), 4.0 ± 4.3 (Han et al., 2020), 5.5 (Gams et al., 2017), 6.66 (range 1 to 41 days) or 8.27 (range 2–69 days) (Storoe et al., 2001), 14.8 (2–81) (In Finland) (Seppänen et al., 2008), again, aligned with the present study (3 days; IQR 2–4). Previous studies list several predictors for length of hospital admission, including diabetes and severity of infections (Gams et al., 2017).

The need for treatment in a closely monitored environment is aligned with the severity of the presentations (Han et al., 2020). Mean duration of intensive care unit admission has been presented using different units, in hours 39.12 (Liau et al., 2018), 38.5 ± 60.4 (Han et al., 2020), or days 6.2 (2–19) (Seppänen et al., 2008), 4.7 (Gams et al., 2017), and in the present study the median of duration was 1 day (IQR 1–2). The need for general anaesthesia, an extended stay admission, and intensive care unit engenders considerable costs which are mostly preventable (Ahmad et al., 2013). The crucial role of social‐level outcomes, including effects on public health, cost‐effectiveness, and allocation of resources, has been highlighted recently (Wrangstål et al., 2024).

The present study has limitations. These include the retrospective nature that may be reflected in the accuracy of the data or missing documentation, however, the records included contained the expected information, and data collection accuracy was validated. In addition, cross‐sectional studies are observational study designs that investigate exposure and outcome as a snapshot in time and can help infer relationships between a disease and other variables, thus revealing associations rather than causations (Arraj et al., 2019). These limitations are shared with several comparable studies globally (Gams et al., 2017; Shah et al., 2013; Ullah et al., 2023). In addition, imaging available varied, and reflected prescriptions by operators at the emergency unit at the general hospital (e.g. panoramic, computed tomography scans). Differences in the diagnostic accuracy for detecting apical periodontitis and other findings, for example, have been previously demonstrated (Estrela et al., 2008), with two‐dimensional imaging potentially not providing the information in toto (Doğramacı et al., 2014). Imaging can vary according to the setting and indication as reported in comparable studies (Bottin et al., 2003; Sánchez et al., 2011). The use of a convenience sample from a single centre has also drawbacks, for example, the findings may not be representative of a wider population and have low power, however, the difficulties in accessing and analysing comparable ‘real life’ datasets should be reiterated. Nonetheless, caries, apical periodontitis, diabetes, and smoking habits are common globally, thus broadening the generalizability of the present study. Records from Aboriginal and Torres Strait Islanders were excluded due to the discordance between the study design (retrospective audit) and the requirements to obtain Ethical approval from the relevant bodies. Diabetes and the smoking habit were the comorbidities considered in the present study as these presented with a higher frequency compared with others, based on a preliminary study. The diabetes type was not discriminated at the data collection stage, and subsequently, in the analysis. The heterogeneity in genetics and clinical presentation of the different types of diabetes has been previously discussed (Cintra et al., 2021; Segura‐Egea et al., 2015). Smoking is a self‐reported variable, thus respondents may suffer from bias, for example, social desirability (Rosenman et al., 2011).

Further comparative multicentre longitudinal studies assessing the association between severe odontogenic infections and the various potential comorbidities are necessary, also considering the potentially limited generalizability of the results of the present study. These should assess systematically endodontic tooth‐related factors to understand possible associations and establish causal relationships. Finally, the paucity of studies in Endodontology encompassing interprofessional clinical practice, such as the management of acute infections in general hospitals, needs to be reiterated.

CONCLUSIONS

Tooth‐related factors, such as molar, mandibular location, untreated caries, and periapical radiolucency were the most common presentations in patients requiring hospital management for severe odontogenic infection. Ludwig's angina was associated with multiple offending teeth and smoking. Diabetes, Ludwig's angina, and submandibular space infections were associated with more complex clinical care.

AUTHOR CONTRIBUTIONS

Laura Petroff: investigation (lead); formal analysis (equal); writing – original draft preparation (lead); funding acquisition (lead). Ruby Richardson: conceptualization (equal); writing – review and editing (equal). Emilija Jensen: conceptualization (equal); Andrew Cheng: supervision. Paul Sambrook: writing – review and editing (equal). Giampiero Rossi‐Fedele: conceptualization (equal); supervision (lead); validation (lead); visualization (lead); writing – review and editing (equal).

ETHICS STATEMENT

The study was approved by the University of Adelaide Human (reference number: H‐2024‐078) and the Central Adelaide Local Health Network (reference number: 17725) Research Committees.

Petroff, L. , Richardson, R. , Jensen, E. , Cheng, A. , Sambrook, P. & Rossi‐Fedele, G. (2025) Association between endodontic, patient‐related factors and severe odontogenic infections; a South Australian retrospective audit. International Endodontic Journal, 58, 727–735. Available from: 10.1111/iej.14204

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request.