High degree of dental neglect characterizes adult patients hospitalized with severe odontogenic infections: a retrospective cross-sectional study

Abstract

Background

Dental neglect is associated with deteriorated oral health, which can lead to the development of severe odontogenic infections (SOI) eventually requiring hospitalization. The clinical impact of risk factors for dental neglect remains unknown. This study aims to identify signs of dental neglect in patients hospitalized with odontogenic infections with the overall goal of early detection and prevention.

Methods

This study is a retrospective cross-sectional analysis of 384 patients previously hospitalized with a SOI. Patients completing the Dental Neglect Scale (DNS) were included. The patients were divided into two groups based on high and low DNS scores, and were compared in terms of clinical, paraclinical, radiographic, and sociodemographic variables using unadjusted and adjusted estimates (chi-square test of independence, independent sample t-test, and binary logistic regression models).

Results

A total of 163 patients completed the DNS (response rate 51.7%), of whom 78 patients (47.8%) had a high DNS score. A high DNS score was significantly related to male sex (p ≤ 0.001), to the indication for having teeth extracted during hospitalization (p = 0.007), and having two or more decayed teeth or teeth with apical periodontitis, with an odds ratio of 3.2 (p = 0.025) and 2.9 (p = 0.019), respectively. The score of Decayed, Missing, and Filled Teeth (DMFT) (p = 0.982) and attendance of dental care prior to hospitalization (p = 0.107) was not correlated with the DNS score.

Conclusions

Approximately half of all patients hospitalized with SOI have a high degree of dental neglect. From a clinical perspective, the study addresses the potential of using the clinical threshold or the DNS in the early identification of patients at risk of SOI. However, further research on the topic is necessary to define the target of preventing patients at risk of SOI within the healthcare system. Exploring dental neglect in a non-selected study population might elucidate the validity and clinical usability of the threshold.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12903-025-06711-x.

Background

Oral diseases represent a global health challenge and reflect social and economic inequalities within and between countries [1]. Marginalized groups and older people are particularly affected by oral diseases, especially in low- and middle-income countries [1]. Ignoring oral diseases such as dental decay, apical periodontitis (AP), periodontitis, and pericoronitis can lead to the development of severe odontogenic infections (SOI) [2].

The term dental neglect was first defined in 1996 as a “failure to take precautions to maintain oral health; failure to obtain needed dental care; and physical neglect of the oral cavity, teeth, and mouth.” [3]. Later, in 2013, “the absence of valuing oral health” was added to the definition [4]. The definition included behavioral ignorance towards dental care, which was explained as a choice to actively ignore care, although the patient had adequate resources and was well-informed [5]. In 2010, the World Health Organization (WHO) described how social determinants account for inequality in exposure and vulnerability to health-compromising conditions in general [6]. This dimension is essential when highlighting health inequalities. Dental neglect has been a topic of interest in pediatric dentistry in the identification of failure-of-care and correct treatment planning [3–5, 7, 8]. To detect neglect of dental health in children, Thomson et al. [3] tested a dental neglect scale (DNS) [3], which was later modified to adults with self-neglect [9]. The DNS has shown high validity and reliability [9, 10].

The majority of patients hospitalized with SOI are characterized by compromised oral health, and it has been hypothesized that dental neglect is a frequent problem in this group [11]. A thorough comprehension of dental neglect as a potential risk factor can be helpful for clinical practitioners in preventing hospitalization with SOI. Early identification of these patients at risk can be used to perform timely and individualized prevention.

Dental neglect has been linked to male gender, patients with a low socioeconomic index [10], and has shown a significant correlation with insufficient dental care behaviors, including lack of interest in seeking treatment for dental decay and trauma, failure to fulfill treatment recommendations [5], seeking dental care for pain relief rather than prophylactic precautions, and in general, a high distrust towards dental services [9, 12]. Clinically significant associations have been identified between dental neglect and a higher overall dental decay experience, tooth loss due to dental decay, increased number and severity of decayed teeth, and increased Decayed, Missing, and Filled Teeth or Surfaces (DMFT or DMFS, respectively) [3, 10, 13]. Poor oral hygiene, accompanied by extensive plaque deposits, has also been associated with dental neglect [9]. Despite these well-documented relations between dental neglect and dental diseases, limited information is available on the impact of dental neglect on the development of more severe odontogenic infections [3, 14]. This study aims to identify parameters useful in the identification and prevention of dental neglect in patients who have been previously hospitalized with SOI. It is hypothesized that patients with a high DNS score would have significantly more teeth with dental decay and AP and report less frequent contact with the dental care system prior to hospitalization compared to patients with a low DNS score. This study aims to determine whether the DNS or a threshold of dental neglect can be used as a tool in clinical practice to identify patients who require a more comprehensive preventive approach. Findings from this study are intended to support clinical practitioners in reducing the risk of SOI and, in the long term, decrease the load on the maxillofacial hospital department caused by SOI admissions.

Methods

This study is a retrospective cross-sectional study that includes a self-reported questionnaire survey, which consisted of the DNS. Risk of bias was examined by applying the “Strengthening the Reporting of observational studies in Epidemiology Statement” (STROBE) checklist for cross-sectional studies [15].

The DNS consists of six items and a 5-point Likert scale (Table 1).

Table 1.

The dental neglect scale

Items	Definitely no				Definitely yes
1. I keep up my home dental care	1	2	3	4	5
2. I receive the dental care I should	1	2	3	4	5
3. I need dental care, but I put it off*	1	2	3	4	5
4. I brush as well as I should	1	2	3	4	5
5. I control snacking between meals as well as I should	1	2	3	4	5
6. I consider my dental health to be important	1	2	3	4	5
Six self-reporting items assessing oral behavior and attitudinal aspects. Each item is given a score 1–5 on a Likert scale ranging from “definitely no” =1 to “definitely yes”=5. *When calculating the score, the item 3 was kept constant

A round-trip translation was performed, and the translation of the DNS into Danish followed the guidelines provided by WHO [3, 16]. A pilot study involving ten interviews with individuals over 18 years was conducted to assess the understanding of the six items from the survey, and the reliability of the translation was evaluated.

Setting

All patients hospitalized with SOI at the Department of Oral and Maxillofacial Surgery, Copenhagen University Hospital, Denmark, between November 2012 and 2019 were contacted. The DNS data were collected from December 2022 to February 2023. A license to conduct the study and handle the data was provided by the Danish Patient Safety Authority (registration: 3-3013-2349/1) and the Capital Region of Denmark Center for Data (registration: P-2019-841). Clinical trial number: not applicable.

Participants

Potentially eligible patients were identified from a cohort of 384 patients hospitalized with SOI. Patients diagnosed with cellulitis or abscess of the mouth by the WHO ICD-10: K12.2 diagnostic criteria were included in the study [17]. The following exclusion criteria were applied: Patients < 18 years of age, deceased patients, patients with infections related to maxillofacial fractures, osteosynthesis material, osteonecrosis of the jaws, and irradiated patients. All eligible patients were contacted by E-boks (the national digital mailbox in Denmark) to complete the DNS survey. Subsequently, non-respondents were contacted by telephone.

Data collection

The DNS questionnaire was self-reported by the participants. For the calculation of the DNS score, all item scores needed to be reversed except item 3, and the total score ranged from 6 to 30, with a high score indicating a high degree of dental neglect [3]. The threshold for a high degree of dental neglect was set at a score of ≥ 13, while a score of ≤ 12 was classified as a low degree of dental neglect [3, 9].

The following data were collected from the hospital records:

• Sociodemographic and health data: Age, sex, comorbidities, smoking and alcohol habits, and drug abuse. The total number of comorbidities was calculated as the sum of all types of comorbidities.

• Clinical and paraclinical data: Suspected odontogenic origin of infection (AP, periodontitis, pericoronitis, recent tooth extraction), location of infection, type of treatment (incision/drainage and/or extraction), number of surgeries, neurosensory complications, necrotizing soft tissue infection or deep neck infection, as well as CRP (C-reactive protein) level and leukocyte count.

• Radiographic data: DMFT score, number of teeth with AP (none, one, two or more), and severity of the periodontal bone loss (≥ 80%, 79 − 66% or < 66%) [18]. Variables were assessed through analysis of panoramic radiographs (OPG). The primary investigator, S.H.D., and RS analyzed the OPGs. In cases of uncertainty or discrepancy, S.S.J. and M.M. were consulted. Only severe dental decay was recorded due to the limitations of OPG projections. The guideline for the radiologic assessment corresponded to the ICDAS system D2 and D3 [19].

• The course of treatment: Symptoms and treatment prior to hospitalization, length of stay at the hospital (LOS), and attendance at follow-up visits.

Study data were collected and managed using REDCap electronic data capture tools hosted at Copenhagen University Hospital. REDCap (Research Electronic Data Capture) is a secure, web-based software platform designed to support data capture for research studies, providing (1) an intuitive interface for validated data capture; (2) audit trails for tracking data manipulation and export procedures; (3) automated export procedures for seamless data downloads to common statistical packages; and (4) procedures for data integration and interoperability with external sources [20, 21]. Some patients completed the survey verbally in cooperation with the principal investigator, who used an interview guide to minimize interview bias.

Study size

The sample size was calculated based on the DMF–T (33%) in the adult Danish population [22]. With 80% power and alpha of 0.05, a minimum of 165 (52.3%) respondents were required to achieve a 95% confidence level and ensure that the actual value was within the ± 5% range.

Statistical methods

Data from REDCap were transferred to IBM SPSS Statistics (Statistical Package for the Social Sciences version 20) for analysis. Descriptive statistics were summarized in frequency tables, and unadjusted data were analyzed using a chi-square test of independence and an independent sample t-test to compare patients exhibiting a high degree of dental neglect with those having a low degree of dental neglect. The data collected were classified into two categories: outcome variables (high or low degree of dental neglect) and predictor variables (oral status based on DMFT and AP). Two binomial logistic regression models were employed to assess the impact of predictor variables on the outcome variables. The models were tested for confounders in terms of age and sex. Initially, the variables were tested for assumptions concerning linearity, multicollinearity, and outliers. The sample size for each independent variable should be greater than 15 cases. Natural log transformations were created for the continuous variables, and combined with logit transformation, these variables were tested for linearity by the Box-Tidwell (1962) procedure. A Bonferroni-corrected alpha level was applied. Outliers were tested by case diagnostics, whereas cases with standardized residuals less than 2.5 were kept in the analysis. The Omnibus test of model coefficient, Hosmer and Lemeshow goodness of fit test, and Nagelkerke R² value were assessed to test the model fit. The odds ratios (95% confidence intervals) were calculated using the binary logistic regression model based on the B-coefficient, and a significant value was used at a cut-off level of p < 0.05. Of the binary logistic regression models, the predicted classification and the actual classification were assessed. The probability of a case classified as “yes” complied with the cut value greater than 0.5. In addition, a Principal Components Analysis, PCA, was conducted to assess the validity of the DNS survey based on the replies. A Varimax rotation with Kaiser Normalization was used for the rotation method, and the rotation converged in 3 iterations before the component loadings were assessed. The total variance of the replies was based on an eigenvalue greater than 1.

Results

Participants, response rate and distribution

The DNS survey was distributed to 315 patients (Fig. 1), of whom 187 responded (59.4%). In total, 24 cases were excluded due to incomplete answers; therefore, the final study population consisted of 163 patients (51.7%). The number of patients with high and low dental neglect was 78 (47.8%) and 85 (52.1%), respectively. The total number of cases in the different variables from the hospital records varied due to data availability (Tables 3 and 4, and 5). Both participants with a high degree of dental neglect and participants with a low degree of dental neglect considered their dental health important (item 6), with 87% and 100%, respectively, answering “yes” or “definitely yes (Table 2).

Fig. 1.

Flowchart of respondents

Table 3.

Sociodemographic characteristics

	Total cohort, n = 384	High DNS, n = 78 (%)	Low DNS, n = 85 (%)	P-value
Age in years,
Mean ± sd	46 ± 17.9	41 ± 14.7	44 ± 16.5	0.193
Sex
Male	182	46 (63.9)	26 (36.1)	< 0.001*
Female	202	32 (35.2)	59 (64.8)
Alcohol
Yes	183	48 (57.1)	36 (42.9)	0.255
No	103	17 (45.9)	20 (54.1)
Missing	98	13	29
Smoking
Yes	130	34 (59.6)	23 (40.4)	0.116
No	162	30 (45.5)	36 (54.5)
Missing	92	14	26
Comorbidity
Yes	186	31 (50.8)	30 (49.2)	0.630
No	198	45 (46.9)	51 (53.1)
Missing	0	2	4
Total number of comorbidities
Mean ± SD	0.7 ± 1	0.8 ± 1.1	0.5 ± 0.9	0.358

Abbreviation: DNS = Dental Neglect Scale, ±SD = standard deviation, n = frequencies, %= Valid percentages noted in parentheses, and *=a statistically significant p-value

Table 4.

Anamnestic information from the hospital records

	Total cohort, n= 384	High DNS, n= 78 (%)	Low DNS, n= 85 (%)	P-value
Symptoms number of days prior to the admission
1 week	198	43 (48.39)	45 (51.1)	0.527
2 weeks	32	8 (66.7)	4 (33.3)
3 weeks	10	2 (50)	2 (50)
1 month+	144	25 (43.1)	33 (56.9)	0.687
Missing	2	0	1
Attendance to control visit
Yes	229	51 (49)	53 (54.2)
No	155	27 (45.8)	32 (54.2)
Missing	0	0	0
Treatment prior to admission
None/none-dentist	252	53 (52.5)	48 (47.5)	0.107
Dentist	111	21 (38.9)	33 (61.1)
Missing	21	4	4
LOS
mean ± SD	3.9 ± 4.45	4.5 ± 5.99	3.6 ± 4.29	0.269

Patient with missing values excluded from the table

Abbreviation: DNS = Dental Neglect Scale, LOS = Len Length of Stay ±SD = standard deviation, n = frequencies, %= valid percentages noted in parentheses, and *=a statistically significant p-value

Table 5.

Clinical, paraclinical, and radiological characteristics from the hospital records

	Total cohort, n = 384	High DNS, n = 78 (%)	Low DNS, n = 85 (%)	P-value
Apical periodontitis, n
None	129	26 (45.6)	31 (54.4)	0.006*
One	79	12 (33.3)	24 (66.7)
Two or more	108	33 (67.3)	16 (32.7)
Missing	68	7	14
DMFTscore
Mean, ± sd	10.77 ± 7.4	11.11 ± 6.4	11.01 ± 7.4	0.880
Decayed teeth, n
None	212	38 (41.3)	54 (58.7)	0.025*
One	40	16 (61.5)	10 (38.5)
Two or more	66	17 (70.8)	7 (29.2)
Missing	66	7	14	0.003*
Mean ± sd	0.98 ± 2.2	1.2 ± 2.095	0.39 ± 0.870
Missing teeth, n,
mean ± sd	3.3 ± 3.5	3.03 ± 2.898	3.45 ± 3.426	0.429
Filled teeth, n,
Mean ± sd	6.5 ± 5.4	6.89 ± 4.747	7.17 ± 5.545	0.746
Mean remaining bone
≥ 80%	233	47 (47.5)	52 (52.5)	0.640
79 − 66%	49	13 (54.2)	11 (45.8)
< 66%	77	11 (57.9)	8 (42.1)
Focus of infection
Apical periodontitis	180	33 (47.8)	36 (52.2)	0.614
Post extraction	99	24 (45.3)	29 (54.7)
Pericoronitis	23	3 (33.3)	6 (66.7)
Other	82	18 (56.3)	14 (43.8)
Type of treatment
Incision and drainage	105	58 (50.9)	56 (49.1)	0.238
Extraction	162	41 (60.3)	27 (39.7)	0.007*
CRP, mean ± sd
1st	133 ± 105	140 ± 114.4	128 ± 90.7	0.468
Leukocytes, mean ± sd
1st	13.6 ± 5.1	13.9 ± 5	12.7 ± 4.6	0.158

Abbreviation: DNS = Dental Neglect Scale, ±SD = standard deviation, n = frequencies, %= valid percentages noted in parentheses, crp = c-reactive protein, and *=a statistically significant p-value

Table 2.

The distribution of replies

Items	Definitely no 1	2	3	4	Definitely yes 5
1. I keep up my home dental care
High	3 (3.8)	5 (6.4)	10 (12.8)	42 (53.8)	18 (23.1)
Low	0 (0)	0 (0)	2 (2.3)	21 (24.7)	62 (76.5)
2. I receive the dental care I should
High	19 (24.4)	18 (23.1)	12 (15.4)	23 (29.5)	6 (7.6)
Low	0 (0)	1 (1.1)	2 (2.4)	17 (20)	65 (76.5)
3. I need dental care, but I put it off
High	10 (12.8)	16 (20.5)	8 (10.3)	29 (37.2)	15 (19.2)
Low	60 (70.5)	18 (21.2)	5 (5.9)	1 (1.2)	1 (1.2)
4. I brush as well as I should
High	3 (3.9)	6 (7.7)	11 (14.1)	27 (34.6)	31 (39.7)
Low	0 (0)	1 (1.1)	1 (1.1)	24 (28.3)	59 (69.5)
5. I control snacking between meals as well as I should
High	17 (21.7)	25 (32.1)	18 (23.1)	13 (16.7)	5 (6.4)
Low	8 (9.4)	11 (12.9)	29 (34.1)	24 (28.3)	13 (15.3)
6. I consider my dental health to be important
High	2 (2.6)	3 (3.8)	5 (6.4)	28 (35.9)	40 (51.3)
Low	0 (0)	0 (0)	0 (0)	13 (15.3)	72 (84.7)

Abbreviation: Frequencies, n (%), in responses to the DNS, and the distribution in percentage noted within a high degree of dental neglect n = 78 (100), and low degree of dental neglect n = 85 (100)

Characteristics of study participants in unadjusted estimates (Table 3)

The mean age of the high and low dental neglect groups was 41 years (SD ± 14.7) and 44 years (SD.

± 16.5), respectively. There was a statistically significant difference in sex distribution between the two groups, with females being more often represented in the low dental neglect group and males in the high dental neglect group (p < 0.001). There was no significant difference in the distribution of smokers and non-smokers (p = 0.116) and for alcohol consumption (p = 0.255) between the groups. Additionally, no differences were observed between the groups regarding the presence and number of comorbidities. Substance abuse and type of disease were excluded due to the study size, except for mental disorder and hypertension, which nonetheless did not differ significantly (Table 3).

The high and low dental neglect groups did not differ significantly in terms of anamnestic data. No difference was found in attendance at control visits at the hospital, nor in symptoms or treatment prior to hospitalization. The LOS was not significantly different either (p = 0.269) (Table 4).

The radiological evaluation revealed a statistically significant difference in the number of AP (p = 0.006) in terms of none or one tooth, with AP being more frequently present in the low dental neglect group. In contrast, two or more teeth with AP were more frequently observed in the high dental neglect group (Table 5). The DMFT score was not significantly different between the groups, with a mean DMFT score of ≈ 11 (p = 0.982). However, analysis of dental decay as a separate measure showed a significant difference with a mean number of decayed teeth of 1.2 ± 2.095 in the high dental neglect group and 0.39 ± 0.870 in the low dental neglect group (p = 0.003). By grouping the continuous variable, there was a statistical difference in the number of decayed teeth (p = 0.025). The low dental neglect group had more often no decayed teeth, whereas the frequency of having one, two, or more decayed teeth was more pronounced in the high dental neglect group.

According to the type of treatment at the hospital, extractions were more frequently indicated in the high dental neglect group (p = 0.007). The two groups did not differ significantly regarding the number of missing and filled teeth, periodontal status, focus of infection, mean CRP, or leukocyte count. There was no difference in the location of infection, the number of surgeries, nor in developing necrotizing soft tissue infection or deep neck infection (Table 5).

Predictors for dental neglect

The assumptions of the binary logistic regression models were met (Omnibus test of model coefficient: p < 0.002). Including the filled, missed, and decayed teeth with the covariates, age and sex, provided 17.7% of the variance in the binary outcome (Nagelkerke R² value), whereas 21.5% of the variance was provided when including the number of teeth with AP. The DMFT score was excluded due to multicollinearity. The outcome revealed that three factors were significantly different between the dental neglect groups: sex (p < 0.001), having two or more decayed teeth (p = 0.025), and having two or more AP (p = 0.019) (Tables 6 and 7). The odds of having a high DNS score when being male was 3.3–3.5 (C.I.=1.594–6.986 and C.I = 1.666–7.192, respectively) as opposed to being female. The odds of the outcome of a high degree of dental neglect were 3.2 times higher (C.I.=1.159–8.721) for patients who had two or more decayed teeth compared to patients without decayed teeth. Likewise, having two or more AP associated with 2.9 times higher odds (C.I.=1.191–6.910) of having a high DNS score than patients without AP.

Table 6.

The effect of the number of decayed, missed, or filled teeth

				95% CI
	B coefficient	P-value	Odds Ratio, Exp(B)	Lower	Upper
Age	−0.015	0.271	0.985	0.959	1.012
Sex; male	1.205	0.001*	3.337	1.594	6.986
Missed teeth	−0.003	0.965	0.997	0.881	1.129
Filled teeth	0.018	0.645	1.018	0.944	1.098
Number of decayed teeth:
One	0.471	0.340	1.601	0.609	4.209
Two or more	1.157	0.025*	3.180	1.159	8.721
Abbreviation: B-coefficient, p-value based on Walds test, Odds Ratio, and its confidence interval (C.I.). Variables entered were age, sex, and the number of misseg, filled, and decayed teeth

Table 7.

The effect of the number of AP

				95% CI
	B coefficient	P-value	Odds Ratio, Exp(B)	Lower	Upper
Age	−0.021	0.109	0.980	0.955	1.005
Sex; male	1.242	< 0.001*	3.462	1.666	7.192
Number of AP:
One	−0.416	0.385	0.660	0.258	1.686
Two or more	1.054	0.019*	2.868	1.191	6.910

Abbreviation: B-coefficient, p-value based on Walds test, Odds Ratio, and its confidence interval (C.I.). Variables entered were age, sex, and the number of AP.

Category prediction: sensitivity and specificity

The dependent variable would correctly classify 50% of the cases at the baseline analysis. Adding the predictor variables to the model, the percentage accuracy in classification in Table 6 was 66.2% and in Table 7 it was 67.6%. In Table 6, the sensitivity was 65.8%, and the specificity was 66.7%, while the positive predictive value was 67.6% and the negative predictive value was 64.8%. In Table 7t, the sensitivity was 67.6%, the specificity was 67.6%, and the positive and negative predictive value was 67.6%.

Principal components analysis

The assumptions of the PCA were approved, and the test results were displayed (Table 8). The test dimensions two factors, high DNS (factor 1) and low DNS (factor 2), respectively. The two factors accounted for 62.6% of the total variance: 40.2% and 22.4%, respectively. Tooth brushing and considering dental health as important loaded strongly to high DNS. The use of professional dental care and postponing treatment loaded strongly on low DNS. Home dental care loaded almost equally to factor 1 and 2, while eating habits loaded weakly on both components.

Table 8.

Outcome of PCA

Items from the DNS survey	Factors
	1	2
1. Home dental care	0.676	0.499
2. Professional dental care	0.187	0.872
3. Postponing	0.005	0.890
4. Tooth brushing	0.878	0.036
5. Snacking	0.197	0.112
6. Importance of dental health	0.802	−0.035

Abbreviation: The influence from items to the two factors. The highest values are highlighted

Discussion

A high DNS score in patients previously hospitalized with SOI was associated with a male sex, multiple odontogenic pathologic conditions, and the need for a more aggressive surgical treatment during hospitalization.

The patients with a high DNS score had significantly more teeth with dental decay and more AP, and the threshold was having two or more decayed teeth or AP, determined by analysis of OPG. Since no differences were detected when having one decayed tooth or one AP, it could be argued that two or more might be the clinical and radiographical threshold for dental neglect. The prevalence of decayed teeth in the Danish adult population has previously been reported as relatively low, with a mean of 0.9 to 1.5 decayed surfaces per individual among those aged 35 to 74 years [22]. In Denmark, apical periodontitis (AP) is present in 3.4% of all teeth and has been identified in every second root-filled tooth [23, 24]. Applying a threshold of two or more AP lesions may lead to the misclassification of patients with root-filled teeth as exhibiting a higher degree of dental neglect, which may not accurately reflect their oral health behavior.

Thus, this study supports the evidence of the association of dental decay with a high degree of dental neglect [3–5, 8–10, 12, 13]. The clinical implication of the present results is that a clinical and radiographic examination could indicate if a patient is suffering from dental neglect. Dental professionals should increase awareness of dental neglect when two or more oral infectious foci occur. However, the determination of dental neglect based on sex, dental decay or AP needs to be assessed in its entirety and exceptions from these characteristics can occur. A systematic review on child dental neglect [5], concludes that it would be a vast simplification to use the number of decayed teeth as the only diagnostic criterion for dental neglect. Our study supports the assumption of oversimplification since a portion of teeth with AP are caused by dental decay and that the number of dental decay and AP explained a low percentage of variance in the outcome of high or low dental neglect. In addition, the sensitivity analysis revealed that the variables only had a mediocre potential to predict the outcome of dental neglect. Therefore, our study acknowledges that more variables should be included to create a diagnostic tool. The approach of this study was that decay and AP increased the odds of dental neglect, but it might as well be a reversed effect: Dental neglect increasing the risk of decay and AP.

No significant difference was found in the treatment pattern prior to hospitalization, which contrasts with earlier findings [5]. The dental attendance prior to the hospitalization included both acute and general dental practitioners and did not distinguish between regular check-ups or emergency appointments as other studies [3, 9, 13]. It might have affected the outcome, since other studies found a significantly different use of professional dental care, whereas dental neglect patients were more prone to use an emergency treatment service rather than prophylactic treatment approaches [5, 9, 25]. If dental neglect mainly is present in dental emergency practice, the effect of identification would be most valuable to implement here. Questions 2 and 3 in DNS regarding the use of and postponing professional dental care were important for the outcome of high neglect according to the principal component analysis, and perhaps these variables should be included as predictor variables of dental neglect.

Male patients were more disposed to exhibit dental neglect, which corresponds with previous studies on neglect [10, 13] and on the use of health care services in general [26]. Dental extractions during hospitalization were more frequent for the high dental neglect group, which may imply more teeth could not be preserved among patients suffering from dental neglect. This finding was expected considering the definition of dental neglect as failure to obtain needed dental care [3]. However, the number of missing teeth between patients with high and low DNS score did not differ, contrary to a study in young adults [10]. In terms of the LOS, the high dental neglect group was on average hospitalized one day longer than the low dental neglect group. The difference was not significant; however the outcome is of clinical relevance considering the resources spent on hospitalization. No other clinical measures pointed towards a more severe course of hospitalization in the group of patients with high dental neglect. The periodontal status and DMFT score did not differ between the groups, which was also contradictory to former observations [3, 5, 13]. It could be argued that the DMFT score, including filled teeth (F), is inconvenient as a measure of objectifying dental neglect since the F component, to some extent, expresses that the patient has requested relevant treatment for their oral pathological conditions. Only a few studies have investigated the relation of age on dental neglect and indicated an accumulation of dental neglect in young adults [12] and elderly [27]. However, in the present study, the high dental neglect group did not differ in terms of patient age.

A critical aspect of the definition of dental neglect was the patient’s choice to actively ignore dental care despite being well-informed and financially capable [5]. It was a challenge to separate these components clinically, and the DNS does not consider this distinction. According to our results, valuing oral health as necessary was almost equally distributed between the two groups, which questions the definition of dental neglect. For the present comprehension of dental neglect, it is essential that what seems to be dental neglect can be founded in other barriers towards dental care. Therefore, the theoretical term of dental neglect needs to be reconsidered to include the effect of social determinants and causalities of neglectful behavior. Prospectively, it is interesting if qualitative research would substantiate the detection of dental neglect [28]. There was potential in using the DNS as a self-assessment tool, and to apply preventive strategies towards patients with increased risk of developing SOI. The DNS survey has several advantages, such as being easy and fast to complete, simple to analyze, none-invasive, and of low-cost. The major disadvantage of the DNS was the imprecision in its measure and lack of test-retest reliability [3, 9]. Furthermore, a risk remained of misconception with self-reporting surveys by obtaining a high score based on high self-criticism and vice versa. The PCA confirmed the 2-components, which resemble prior findings of dental neglect [3, 9]. However, according to the pilot interviews, recommendations on tooth brushing and snack control were not common knowledge, which the PCA confirmed.

Since dental neglect has been associated with a lack of interest in dental services [5, 9, 10, 13] a final study size of 51.7% can be considered a high response rate. However, the calculated sample size based on DMFT score [28], demanded a minimum of 52.3% to obtain generalizability. The difference was small why the underpowered study might increase the disparity between the two groups in terms of larger confidence intervals, overlapping SD and an increased risk of type 2 errors emerging. The comparison of the high and low would presumably be more pronounced or even different if the patients hospitalized were compared to a random sample of the general population.

The two main limitations of the study are the retrospective design and the single-center setup. The single-center setup of this study may increase the risk of various types of bias and compromise its generalizability. SOI is a global burden on maxillofacial hospital clinics, and it is recommended that future studies focus on a multicenter study design to increase the generalizability of the results. The time between hospitalization and the completed DNS might affect the results, and assumably, the high DNS group would be underestimated due to lack of interest in participation or possibly behavioral changes after admission. Both AP and dental decay was also most likely underestimated, since only high degrees of mineral loss could be registered using OPGs. The wide coverage of OPGs often comes at the expense of detailed resolution. Therefore, early signs of AP, such as minor changes in the periapical tissues, may be too subtle to be detected on these images. The amount of data was a strength of the study.

A dropout analysis (Additional file 1) revealed that there was no significant difference between the respondents and non-respondents, suggesting that the patients who completed the DNS survey were representative of the patients hospitalized. In a study among Norwegian adults, the prevalence of high dental neglect was 20% in a random sample (n = 2000) from a national register [25]. In contrast 52.1% of the hospitalized patients in the present study had a high degree of dental neglect, which supports the presumption of the patients hospitalized are more vulnerable to dental neglect. In the cross-sectional study design no information of causality is obtained, which could have been of interest, however it was advantageous to chart the prevalence of dental neglect to elucidate the challenge in the dental health sector.

There is considerable potential to reduce the high-cost specialist task through early prophylactic efforts [29]. Further research is needed to elucidate the best way of handling the patients with DN whether in need of more adaptation to the dental care system, a more radical treatment approach, or screening of dental neglect in pediatric dentistry or identifying dental needs in other healthcare sectors. To reduce the challenge of dental neglect, introducing preventive strategies in terms of systematic recruitment to the dental care targeting the social determinants of health inequalities could be a solution. For instance, a screening of dental neglect at the job center, at the general practitioner or at nursing homes. When keeping social determinants in mind [6], the most optimal preventive manner is to target the issue in a socioeconomic and political context to increase the positive feedback mechanism on structural determinants with the goal to decrease health inequality.

Conclusion

A high degree of dental neglect was associated with the clinical expression of multiple odontogenic pathological conditions. Having two or more teeth with AP and/or decayed teeth was associated with having a high DNS score. Thereby, the hypothesis that patients with a high degree of dental neglect have significantly more teeth with dental decay and more teeth with AP could be confirmed. In addition, being male or in the need for tooth extraction during hospitalization was significantly associated with obtaining a high DNS score. The hypothesis that patients without contact with a dentist prior to hospitalization exhibited a high degree of dental neglect compared to patients referred by a dentist was rejected. In conclusion, the theoretical term dental neglect can potentially guide oral health initiatives, yet further research on this topic is necessary. Our study acknowledges that more variables should be included to create a diagnostic tool, yet the threshold has potential to identify patients at risk of SOI and could be a guideline for the dentist to explore causalities of neglectful behavior by the patient. Further research is encouraged to elucidate which variables other than AP and dental decay that may additionally contribute to refining a DNS threshold.

Abbreviations

AP

Apical periodontitis

CI

Confidence Interval

CRP

C–reactive protein

DNS

Dental Neglect Scale

DMFT

Decayed, Missing, and Filled Teeth

DMFS

Decayed, Missing, and Filled Surfaces

ICD

10–International Classification of Diseases, 10. revision

ICDAS

International Caries Detection and Assessment System

LOS

Length of Stay

OPG

Orthopantomogram

PCA

Principal Components Analysis

REDCap

Research Electronic Data Capture

SD

Standard Deviation

SOI

Severe Odontogenic Infections

SPSS

Statistical Package for the Social Sciences

STROBE

Strengthening the Reporting of Observational Studies in Epidemiology

WHO

World Health Organization

Authors’ contributions

S.H.D. contributed with conceptualization, methodology, data curation, formal analysis, investigation, and writing – original draft, R.S. contributed with conceptualization, data curation, formal analysis, and writing – review & editing, S.S.J. contributed with conceptualization, methodology, supervision, and writing – review & editing, E.B.Ø. contributed with conceptualization, supervision, writing – review & editing, and M.M. contributed with funding acquisition, conceptualization, methodology, supervision, formal analysis, and writing– review & editing.

Funding

Open access funding provided by Copenhagen University. The Research Foundation of the Danish Dental Association supported the research project.

Data availability

Due to the sensitive nature of the clinical data, the anonymised datasets supporting the findings of this study are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

A license to perform the study and to handle the data was provided by Danish Patient Safety Authority (registration: 3-3013-2349/1), and Capital Region of Denmark Center of Data (registration: P-2019-841). In accordance with Sect. 46(2) of the Danish Health Care Act, informed consent for handling anonymized patient data was waived. This study was performed in compliance with the ethical principles outlined in the Declaration of Helsinki.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.