Understanding Oral Self-Care Practices Among People with Diabetes—A Qualitative Study

Abstract

Background: A bidirectional association between diabetes and oral health is well established, yet oral self-care is overlooked in diabetes management. Health Belief Model (HBM)-guided oral care interventions have exhibited promising outcomes in the literature but have not been used to guide oral self-care interventions designed for people with diabetes (PWD). Positioned at the early conceptualization and design stage of such a program, this developmental study was to identify self-perceived needs in oral self-care practices and to obtain preliminary feedback among PWD about the blueprint of a new program—DiaOral^©. Methods: We conducted semi-structured interviews with 15 PWD recruited from a large healthcare system, with a goal to recruit patients from racially/ethnically diverse urban/suburban zip codes. Interviews explored participants’ oral self-care practices in relation to diabetes. Sample DiaOral^© content and images on a blueprint were presented and feedback was solicited. Braun and Clarke’s reflexive thematic analysis was used to code and interpret transcripts, aligning emerging themes with HBM constructs through team-based consensus. Results: Three major themes and 27 sub-themes emerged: (1) lack of knowledge on optimal oral care, (2) low perceived importance of preventive care and oral health in diabetes, and (3) low self-efficacy for performing effective oral self-care. Participants expressed satisfaction with the content and their perceived confidence and interest potentially in using the DiaOral^© program based on their preliminary review of the blueprint. Conclusions: Findings support the relevance of HBM constructs in shaping oral self-care among PWD. This developmental study suggests that the DiaOral^© blueprint is ready to move forward to website prototype development. Future work will finalize the program and evaluate its efficacy among PWD.

1. Background

In the United States, diabetes and periodontal diseases are highly prevalent chronic diseases, both having a pronounced impact on racially and socioeconomically disadvantaged populations [1,2,3]. Oral problems are more prominent among PWD than those without diabetes. More than half of adults with diabetes have periodontal disease [4] and about one-third have untreated caries [5]. PWD, especially those who are racially or socioeconomically disadvantaged, experience greater oral health disparities [6]. Periodontitis is an overlooked but highly prevalent diabetes complication that not only impacts a person’s ability to chew or speak, but also poorly controlled periodontitis may contribute to insulin resistance and reduced efficacy of diabetes medications [7,8,9]. Subsequently, untreated periodontal disease may jeopardize glycemic control and increase the risk for developing more severe diabetes complications, such as cardiovascular and kidney disease [10,11,12,13].

Despite the severity of these complications, oral diseases are largely preventable through optimal oral self-care and preventive care measures. Additionally, dental treatments for periodontal disease and practicing oral self-care, such as flossing, may help reduce blood glucose [14,15,16,17]. However, individuals with diabetes are less likely to practice effective oral self-care and often delay treatment until oral diseases become more severe and costly to manage [4,14,15]. This delay contributes to a detrimental cycle in which worsening oral health further complicates diabetes management, leading to increased risk of systemic complications. Given that oral diseases are largely preventable, there is a critical need to develop interventions that promote better oral self-care behaviors and improve preventive dental care among individuals with diabetes.

One promising approach to improving oral self-care behaviors is through the application of the Health Belief Model (HBM) [16]. HBM-based interventions have demonstrated preliminary efficacy in promoting oral health behaviors among various populations, such as children and older adults [17,18,19]. However, there is a notable gap in research exploring the effectiveness of HBM-based interventions specifically among individuals with diabetes. Addressing this gap is essential to developing specific diabetes self-management strategies that assist PWD to improve oral health and glycemic control and prevent diabetes complications. To this end, we developed the DiaOral^© (digital technology-assisted oral health education and self-care interventions) program with the vision that in the future, such a program can be readily used as a digital educational and intervention administered at primary care for PWD.

Informed by the HBM, this qualitative study was designed to identify self-perceived needs in oral self-care practices to inform the content and structure development of the DiaOral^© program, and to obtain preliminary patient feedback about the blueprint of the DiaOral^© program.

2. Methods

In this qualitative narrative study, we conducted 15 semi-structured interviews with a small group of diverse patients with diabetes from the University of Cincinnati (UC) Health System, a large healthcare system in Cincinnati, Ohio. Patients’ electronic medical records were reviewed and extracted by designated medical informatics staff according to the inclusion criteria in the study protocol. To recruit a small group of diverse adults from the greater Cincinnati area, we invited patients who resided in the top five most diverse zip codes in urban areas and the top five most racially diverse zip codes in suburban areas, based on statistics retrieved from the City of Cincinnati Health Department and the American Community Survey [20,21]. The study sample size was anticipated to be around 10 to 20. According to Creswell, generic qualitative study with thematic analysis requires 5 to 25 participants. [22] However, a newer work by Guest et al. indicates 92% of themes emerged within the first 12 interviews, with additional interviewers producing minimal new information. [23]

The inclusion criteria of this study were as follows: (1) ages 18 years or older, (2) visited UC Health clinics in the past year, (3) diagnosis of type 2 diabetes mellitus (T2DM) or type 1 diabetes mellitus (T1DM), (4) HbA1c ≥ 6.5% in the most recent test, and (5) able to read and speak English. People were excluded if they had any of the following conditions: (1) pregnancy, (2) edentulous, (3) ever had a diagnosis of end-stage renal disease, heart failure, stroke, blindness, dementia, Parkinson’s disease, or other severe diseases and conditions that may interfere with their participation in this research study.

Potential participants for this study were identified in the UC Health Epic EHR system (Epic Systems Corporation, Verona, WI, USA). We extracted information on eligibility and obtained contact information for potential participants. One-time unsolicited email invitations were sent to those who might be eligible for the study. People who decided to participate and signed the informed consent sheet were asked to fill out a screening questionnaire to confirm their eligibility, obtain sociodemographic data and determine availability for a one-hour interview. Study data were collected and managed using REDCap electronic data capture tools (version 16.0.27) [24]. A semi-structured interview was conducted by trained research associates. All interviews were conducted virtually using Teams, Zoom, or Google Meet, depending on the participant’s preference. Following an interview guide listing leading questions, the interviewer conducted the interview with follow-up prompts as needed to inquire about the participants’ oral health and diabetes management beliefs, behaviors and perceptions about the relationship between oral self-care and diabetes management. As the interview proceeded, off-script exploration or clarification prompts and questions were used to elicit more details while remaining focused on the targeted topic. In a final set of questions, an example of the DiaOral^© working draft containing pictures of DiaOral^© content and illustrations of its website figures was shown to participants. Participants’ feedback and comments about the DiaOral^© blueprint were collected. All the interviews were recorded, and transcripts were clarified and entered into QDA Miner Version 2024.0.5 (Provalis Research, Montreal, QC, USA) for data analysis.

Braun and Clarke’s approach to reflexive thematic analysis (RTA) was used to code and analyze the interview data. RTA is a purely qualitative approach that prioritized the researcher’s subjectivity as an analytical tool. Both semantic and latent coding methods were utilized to best capture the rich information presented in the dataset [23,24]. Therefore, as opposed to using a codebook, the data were open-coded initially and adjusted later in a recursive manner as the researchers familiarized themselves with the full dataset. In this process, discrepancies between coders were discussed as a team to reach consensus. Such an approach was adopted because, according to Braun and Clarke, this approach not only describes the information items residing in the dataset ascribed to respondents’ perceptions but also allows the creative identification of hidden meaning/relationships/assumptions, and theoretical meaningfulness that are underpinning the information items, as appreciated in researchers’ perceptions to the phenomenon and theory [24]. Interpretation of the initial codes involved a collaborative sense-checking of ideas to explore multiple interpretations using the HBM as the guide. We did not seek intercoder reliability or consensus, as RTA prioritizes the researcher’s subjectivity as an analytical tool [23,24,25]. The study did not adhere to strict epistemological traditions, such as interpretative phenomenological analysis.

At the theme-generating phase, informed by the HBM, we used a combination of deductive and inductive approaches [26]. A deductive approach was first utilized to guide the understanding of the code clusters and the sub-themes in relation to the HBM constructs that was used as the theoretical guidance (Figure 1). An inductive approach was then utilized to discover major themes that resided within the dataset across HBM constructs and represented the most meaningful problems and needs in PWD’s oral health perception and self-care practices. The feedback about the perceived accessibility and acceptability of the DiaOral^© blueprint was summarized.

Two researchers (K.B. & M.T.) conducted the interviews and pre-coded the transcripts. A third coder (Y.Z), who holds the knowledge as an “insider” of the topic area, open-coded the transcripts and conducted the initial data analysis. Peer debriefing was carried out twice a week initially then switched to weekly after all the coders (K.B., M.T., and Y.Z) familiarized themselves with the dataset. Y.Z. and S. G. L. hold the knowledge and years of research experience with oral health related to diabetes, engaged collaboratively in the reflexive process in the data analysis and met weekly to review the initial open-codes and adjusted the code as needed by a co-coding process. The reflexive memo was documented as notes in the QDA Miner Version 2024.0.5 (Provalis Research, Montreal, QC, USA). Coding and thematic analysis history were saved as QDA files with dates indicating the version of the files. In accordance with the principles of RTA, we recognize that the researcher is an active agent in knowledge production and that our subjectivity is a primary analytical tool [26,27,28].

Throughout the analysis, we practiced collaborative reflexivity. Rather than seeking inter-rater reliability or consensus in a positivistic sense, we used discussions to sense-check interpretations and explore multiple assumptions. While the HBM provided a deductive guide for theoretical meaningfulness, we remained reflexive about how this framework prioritized certain beliefs (e.g., self-efficacy and barriers) over others. This statement serves to acknowledge that a different research team, operating from a different paradigmatic or clinical position, might have construed the relationships between these codes and themes differently.

Basic descriptive analyses with means and percentages summarized the sociodemographic characteristics of the participants and their oral self-care behaviors (i.e., tooth brushing, flossing, use of mouthwash, types of oral hygiene tools used, etc.).

SPSS Version 29 (IBM Corp., Armonk, NY, USA) was used for analyzing quantitative data, and QDA Miner Version 2024.0.5 (Provalis Research, Montreal, QC, USA) was used for analyzing qualitative data. The Standards for Reporting Qualitative Research (SRQR) were used to guide the data reporting process of the study. The SRQR checklist was used to ensure that details in study design, data collection, and data analysis were described in this report [29]. The University of Cincinnati IRB reviewed the study documents and approved the study.

3. Results

3.1. Sample

Of the 30 people who responded to the emailed invitations, 15 people were enrolled and completed the interview. Among the other 15, 1 chose not to participate after reading the informed consent information, 3 were excluded due to being edentulous or missing data on eligibility, 5 did not respond to the follow-up call/email to schedule an interview, 5 did not attend the interview, and 1 responded several months after the study ended.

3.2. Sociodemographic Characteristics

Among the 15 participants, including 9 women and 6 men, there were 7 Black, 6 White, 1 American Indian or Alaska Native and 1 who preferred not to disclose race and ethnicity. Thirteen had T2DM and 2 had T1DM. The average age was 59.4 years with an age range from 28 to 87 years. The average years since DM diagnosis were 15.9 years (range 1–37 years). Eleven participants were from diverse suburban areas and four were from urban areas. Nine participants were college graduates and the other six attended college. Seven participants had private insurance, six had Medicare and two had Medicaid (

Table 1. Participant sociodemographic characteristics (n = 15).

Characteristic	n (%)
Age, Mean (SD) Min–Max	59.4 (17.3), 28–87
Gender Female Male	9 (60.0) 6 (40.0)
Race American Indian or Alaska Native Black White Prefer not to say	1 (6.7) 7 (46.7) 6 (40.0) 1 (6.7)
Education Level College graduate or above Some college or Associate Degree	9 (60.0) 6 (40.0)
Marital Status Married or living with a partner Single/Divorced/Separated/Widowed	9 (60.0) 6 (40.0)
Insurance Employer/Private Insurance Medicaid Medicare	7 (46.7) 2 (13.3) 6 (40.0)
Years of DM a, Mean (SD), Min-Max	15.9 (12.7), 1–37
Type of Diabetes T1DM T2DM	2 (13.3) 13 (86.7)
Residence Urban Suburb	4 (26.7) 11 (73.3)

Note. ^a: Missing data n = 1, DM: Diabetes Mellitus; T1DM: Type 1 diabetes; T2DM: Type 2 diabetes.

).

3.3. Oral Hygiene Behaviors

Regarding oral self-care practice and preventive care, seven participants reported they consistently brush twice daily and eight reported brushing less often than twice daily. Most reported brushing teeth at least 2 min each time (13 out of 15), using fluoride toothpaste consistently (11 out of 15), and using mouthwash from time to time as a part of their oral hygiene regimen (10 out of 15). However, most participants reported that daily flossing is not part of their oral hygiene practice (9 out of 15). In addition, among those who reported daily flossing, two were using a water flosser only, two were using toothpicks/floss picks everyday as needed at times that may or may not be during the tooth brushing times, and only two were “traditional flossers” who used the string floss every day before or after tooth brushing. Although few used string floss (3 out of 15), eight participants reported they had a habit of using toothpicks or pointy floss picks as a means to remove food debris stuck between teeth. Also, eight participants reported using electronic toothbrushes and eight reported they brush their tongues when brushing their teeth (

Table 2. Oral hygiene behaviors (n = 15).

Oral Hygiene Behaviors	n (%)
Tooth brushing Brush twice a day Less than twice a day	7 (46.7) 8 (53.3)
Brush at least 2 min each time No Yes	2 (13.3) 13 (86.7)
Use fluoride toothpaste No Yes Not sure	1 (6.7) 11 (73.3) 3 (20.0)
Preventive dental visit twice a year No Yes	6 (40.0) 9 (60.0)
Use mouthwash No Yes	5 (33.3) 10 (66.7)
Floss every day No Yes, use string floss Yes, use floss picks Yes, use water flosser	9 (60.0) 2 (13.3) 2 (13.3) 2 (13.4)
Use string floss No Yes	12 (80.0) 3 (20.0)
Use electric toothbrush No Yes	7 (46.7) 8 (53.3)
Use toothpicks/pointy end of floss picks No Yes	7 (46.7) 8 (53.3)
Brush tongue No Yes	7 (46.7) 8 (53.3)

).

3.4. Oral Self-Care Belief and Practice with Diabetes

3.4.1. Codes and Sub-Themes in HBM Construct Categories

We identified a set of sub-themes and codes that were derived from the thematic analysis. Codes and sub-themes were summarized in eight categories with each correlated to one HBM construct, namely: (1) knowledge, (2) health literacy and eHealth literacy, (3) perceived susceptibility and severity to oral disease related to diabetes, (4) perceived benefits of practicing oral self-care and preventive care, (5) perceived barriers of practicing oral self-care and preventive care, (6) perceived self-efficacy of practicing oral self-care and preventive care, (7) oral self-care and preventive care actions, and (8) oral health conditions (Figure 1).

Codes and sub-themes with example quotes in relation to the key HBM constructs were summarized in

Table 3. Thematic analysis results of the semi-structured patients’ interviews based on the Health Belief Model (HBM) (n = 15).

HBM Constructs in DiaOral©	Sub-Themes and Codes	Sample Quotes *
HBM—Modifying Factors
1. Knowledge (M1, M2, M3)	Lack of knowledge of basic oral care routine	- “How long do [I] brush? until, I don’t know, until I think I’ve covered everything that I need to.” (P12)
	Limited understanding of healthy mouth condition	- “[gum bleeding a couple times a week] Um, I mean, I don’t… I guess there’s really nothing certain. Maybe I brush too hard.” (P19)
	Limited understanding of DM-OH interaction    ○ DM affects everything    ○ Unsolved confusion	- “I just think that anything could possibly go wrong with diabetes. “ (P18) - “I know diabetes is linked to an awful lot of things, but as far as the mouth is concerned, I don’t know” (P24)
2. Health Literacy and eHealth Literacy (M1)	Healthy skepticism of website info	- “You know, obviously everybody uses Google. That’s the first thing you go to get articles and sometimes the first articles that pop up are the advertising ones that, you know, they want you to buy something from them. But I don’t know. I don’t know how to tell a reputable source of health information” (P15)
	Recommended by authorities	- “a lot of it come from the dentist and a lot of it comes from Google. But. It always is nice when somebody that’s a healthcare provider is telling me and not Google because I would be a Google doctor If I relied on Google to give me the answers that I needed.” (P16)
	Varied comfort level with online info    ○ Online info seeker    ○ Not tech savvy but can do/keep up with the world    ○ Don’t use the internet for info	- “Normally I’ll Google it for a time of technology, but I know that even with Google everybody’s body is different.” (P17) - “No, I do it because it’s, it is what is in now.” (P9) - “I use internet… But honestly, I really rely on my wife because she knows how to do that. She knows what to look for.” (P29)
HBM—Individual Beliefs
3. Perceived susceptibility to and severity of disease (M1, M2, M3)	Low perceived importance	- “Like the primary thing I’m focused on is weight control and foods. Didn’t even think about teeth. “ (P2)
	Lack concern about poor oral health	- “it doesn’t really affect it. I can eat. It’s just that I know… I like to eat things that may not need to be chewed with upper teeth, like steak, and you know, meat and things like that. I know that puts a lot of pressure on my front teeth.” (P18)
	Barrier—never told by health provider	- “Never. Never. No, I have not had one doctor relate diabetes to the condition of my mouth or my teeth.” (P15)
	Barrier—limited resource to know	- “Like when I was diagnosed with diabetes, you know, I went to Web MD, I went to Mayo Clinic. Those standard types of sources to get quick information. [Interviewer: And? You didn’t happen to see anything about oral health on any of those sites?] I did not, no!” (P2)
4. Perceived benefit of practicing oral self-care and preventive care (M2, M4)	Unaware oral self-care benefit diabetes control	- “I don’t know how your oral health affects your diabetes”(P26)
	Not sure of the value/dentist wants to do extensive treatment	- “it was something that he observed when I went in for something else [for regular preventive care]. The impression I got is that he tended to want to do gum surgery on every person that walked in the office.(15)
	“Maybe” thought about the benefit of good self-care	- “I haven’t got the habit of brushing my teeth at night before I go to bed… I know it would probably… It would probably eliminate the big chunk of the problems that I have all the time with my teeth.” (P15)
	Catching early signs can prevent later issues	- “I typically catch things that pop up because I try to be aware of what’s going on, just because I am a diabetic. I don’t want things to get worse if I could catch them.” (P17)
5. Perceived barriers of practicing oral self-care and preventive care (M4, M5)	Confusion about correct OH strategy	- “after 10 years, the facts change. You know, like eggs are good for you & eggs are not good for you. That type thing. So, you know, like, oh, well, electric toothbrushes are good for you. Oh, no. They do much damage. They do damage…So it’s just like OK, just try if we see what… you know, shoot for the best.” (P14) - “I don’t know whether those [floss picks] are any good compared to flossing yourself with a full piece of floss, you know.” (P15)
	Confusion about choosing proper OH tools	-“I’m just not sure if it [Waterpik] really doing what I needed it to do.” (P29)
	Admits low priority preventive dental care	- “I’m very bad at that. I’d rather go to the doctor. I’m probably opposite than most people. I’d rather go to the doctor than the dentist… So, I don’t go. So, I have not been to the dentist. I hate to say in a year I have not.” (P9)
	Food easily stuck btw teeth/big gaps btw teeth	- “But my teeth now have lots of room which is, which is why, in parts of my mouth that I said, I have gaps. Not pretty, pretty good size. Not tooth width but, as far as a piece of floss is concerned, they’re pretty big.” (P12)
	No dental “home”/no established routine	- “the last time I probably been to the dentist, it may be about. couple years ago.” (29)
6. Perceived self-efficacy of practicing oral self-care and preventive care (M4)	Admit do less than should	- “I don’t brush my teeth every day, even though I know I’m supposed to.”(P26)
	Alerted observer	- “I do try to look at my own gums and look at my own teeth and see if I see anything that’s alarming and try to get that checked out if so.“(P17)
	Shame/regret/judgment	- “I should be drawn and quartered, I have. I’ve had dental insurance for years; I haven’t been to the dentist.” (P14)
	Want to know more/need additional help	- “They tell us how to take care of, you know, our eyes. But like, how is our oral health like? How is that? How is that? You know, how does that all make sense? Like, how is it all grouped together?” (P16) - “I suppose I could enlist the help of my wife to urge me to brush my teeth at night.” (P15)
	Overconfident	- “I think my teeth were fine. I didn’t go to a dentist though”(P26)
HBM—Outcomes
7. Oral self-care and preventive care actions (M4, M5)	Incomplete care    ○ Never brush tongue    ○ No or partial floss/use toothpicks    ○ Not a flosser	- “No, I don’t brush my tongue. I don’t feel like I need to brush my tongue.” (P12); “I mean. I do it [floss] every day, but I don’t do my whole mouth every day. I guess I do spots.”(19);” I do not floss. Never.” (P13)”
	Inconsistent care pattern    ○ Multitask during oral routine    ○ No time/in a rush/tired    ○ Situational based care	- “It depends. Sometimes I will hold it in my mouth till I maybe get halfway dressed…And then sometimes I spit it out before them. It just depends on what I’m doing. Sometimes it’s longer than others. I would say that routine doesn’t have any set time.” (P9); “In a hurry, less than a minute with the mouthwash. And then, when I really take my time, I take at least 5 min”(P14);” try to floss daily, but that doesn’t always happen because life is busy of course.”(P17); “if I got something stuck in my teeth, like if I was eating corn.” (P26)
	Long standing habit never been questioned    ○ Brush 2 times a day is the only oral care routine    ○ Brush less than 2 times a day    ○ Neglect tooth brushing	- “Brush twice a day. That’s pretty much it.” (P2); “I only brush once a day and I don’t floss because my gums in great shape, and my teeth have very little tartar. (P13); “When I do it [brush teeth], I do it right, but I don’t do it every day.”(P12)
	Multiple tools user	- “I try to floss. You know, daily doesn’t always happen. I use the regular floss that they use at the dentist office. I do get the picks. Sometimes I notice that when I get the picks there it’s easier to use because I can pick it and take it with me.”(P17); “I do something three times a day and there’s something could be two times Waterpik, one time brushing or maybe two times brushing one time Waterpik” (P24)
8. Oral Health Conditions (M3)	Not good/Bad    ○ Gum issues    ○ Multiple missing teeth    ○ Untreated/active issue present	- “I’ve had that gum surgery a long time ago. They check my, what do you call it when they puncture your gums to see the depth of it? I have that done on a somewhat regular basis and my previous dentist has not mentioned any concern over the depth of my… the thing that they checked with the thing that sticks in your gums.” (P15); “Lots of missing teeth…Millions of them. Yeah, right in the back molars.” (P24); because I’ve got two teeth missing, that I’m getting implants for, so they’re in the process of healing. “ (P15);” I had a root canal in August. They just did a crown or they did a temporary crown and I’m waiting for the permanent crown. And then I’m gonna start periodontal treatment. like the deep cleaning” (P16)
	In good condition	- “I can’t say I am [concern on my oral health]. I mean, my gums aren’t bleeding right. I would be concerned if I had bleeding gums. I’m pretty much free of pain so.” (P24)

* Participant number indicated by P number. Note: DM = Diabetes mellitus; OH = Oral health; btw = between; DiaOral^© Modules M1 = Know your mouth; M2 = Diabetes and oral health; M3 = Self-screening for oral problems; M4 = Practice optimal oral self-care; M5 = Choose the right tools; Shaded and bolded = HBM constructs in Figure 1.

into three categories: HBM modifying factors, HBM individual beliefs, and HBM outcomes. HBM constructs were not merely present in participants’ narratives but were organically interrelated.

Knowledge and health literacy were the key HBM modifiable factors we examined in the study. They shaped a “functional” perspective among the participants. We found oral self-care was motivated by the immediate ability to eat or speak rather than long-term health maintenance. Oral preventative care is not prioritized because participants normalized symptoms like bleeding or tooth loss as inevitable outcomes of aging or declining of overall health due to diabetes rather than recognizing them as modifiable conditions requiring a proactive, systematic oral self-care regimen. There was a distinct lack of understanding regarding the bidirectional mechanism between periodontal health and glycemic control. Health literacy was characterized by “authority-seeking skepticism”. Participants actively sought online information but distrusted it as commercially biased, preferring direct clinical recommendations. Consequently, knowledge is limited not only by a lack of information access, but also by the absence of integrated, credible messaging from trusted primary care providers. Perceived self-efficacy as a key HBM individual belief was presented as a complex spectrum of “shame versus overconfidence.” Many participants admitted they did “less than they should,” with narratives often steeped in shame or self-judgment regarding their dental history. In contrast, some displayed “unfounded overconfidence,” claiming their teeth were “fine” despite reporting long-standing habits that excluded flossing or the basic twice-daily brushing. As a result, by normalization of symptoms like gum bleeding as “nothing certain,” participants constructed a psychological buffer that reduced the perceived urgency of preventative action (HBM perceived severity).

These belief structures eventually contributed to a “situational based” oral self-care pattern rather than consistent routine actions. Oral hygiene such as flossing was often reactive, i.e., triggered by food getting stuck, and largely interrupted by how rushed/tired in daily life, thus leading to inconsistent and fragmented patterns. The physical oral health conditions reported (e.g., multiple missing teeth) were subsequently perceived by participants as tolerable “fateful realities” because “diabetes affects everything”, rather than the terminal outcome that could have been modified with optimal preventive behaviors.

3.4.2. Major Themes

We used thematic mapping to examine candidate themes and eventually identified three major themes: (1) lack of knowledge about optimal oral care, (2) low perceived importance of oral care in diabetes, and (3) limited self-efficacy on optimal oral self-care. (Figure 2).

The first major theme, lack of knowledge about optimal oral care, encompassed a lack of knowledge of basic oral care routines and having a limited understanding of healthy mouth conditions. For example, some people determined tooth brushing time purely by subjective measures, as “until I think I covered everything” or “brush as long as I feel right.” Some people felt indifferent about gum bleeding while brushing and attributed it to “nothing certain, maybe brushing too hard”. When asked, “how do you usually find out health-related information?” People would mention that they prefer to hear it recommended by professionals. And, in the era of the internet, the participants exhibited varied comfort levels with obtaining information online. Many people were online information seekers, many others were not technology savvy but were capable of using online technology when needed. Very few reported not using the internet to seek online health information if it is not presented to them or they rely on family members to do it for them. Many people mentioned they would Google for information online. For example, they would say, “I Google a lot”, “I’ll Google it for a time of technology”, [paraphrased, I will Google it because now it is the time of technology.] and “I do because it’s what is in now.” No matter if people used or did not use the internet to seek online information, they always preferred to receive recommendations from health professionals. We found that among online information seekers, many people have a healthy skepticism toward web-based health information. They reported, “It always is nice when somebody that’s a healthcare provider is telling me and not Google, because I would be a Google Doctor if I relied on Google,” and “obviously everybody uses Google… but I don’t know how to tell a reputable source of health information.”

The second major theme, low perceived importance of oral care in diabetes was evident in that many participants seemed unaware that oral care is important to diabetes care. A participant reported, “the primary thing I’m focused on is weight control and foods, didn’t even think about teeth” (P2). Some people admitted preventive dental care is a low priority, as long as “I can eat, [then] it doesn’t bother me” (P18). This lack of concern about poor oral health was common among some of the interviewees, who we referred to as indifferent. They never considered that poor oral health was preventable through self-care. Some reported that poor oral health is a natural outcome of aging, regardless of diabetes. And if they had heard about someone else with a worse oral health situation, they would use that as a reference and feel content with their own condition. For example, when asked, “Do you feel like your mouth condition is related to your diabetes or it has anything to do with it?”, a participant reported, “No, I’m just an elder, at this point I know a lot of people don’t have many teeth as I have at 60. Most of the people I know have a bridge or a partial or something.” (P22). Consequently, people managed to find a way to live with their oral health issues. For example, although reporting gum problems with bleeding and having missing teeth, when further asked, “You said the pain was gum-related, what do you usually do about that?”, responses were, “nothing, just deal with it.” (P19) and “Yeah, it’s annoying, but it’s tolerable” (P19), or, “how would I describe the condition of my mouth? Function, functional” (P12). Even when food options were limited due to oral conditions, PWD may still choose to live with these issues despite the challenges, such as, “I have a lot of back teeth that are missing and I’m relying on the ones on the side and the front, for chewing now…I like to eat things that don’t need to be chewed with upper teeth, like steak, meat and things like that. I know that puts a lot of pressure on my front teeth, and I don’t know what to eat.” (P18).

As to the bidirectional relationship between diabetes and oral health, many participants had a vague and fateful belief that diabetes can lead to many health problems in their body, i.e., “everything seems to go hand in hand with your diabetes and everything that’s wrong with you, it seems like it’s your diabetes.” (P19); “I don’t really know much. I just know that, like diabetes, can make a lot of things fail and deteriorate.” (P26); “Everything related to diabetes (P3)”, “I just think that anything could possibly go wrong with diabetes.” (P18). Respondents seldom articulated that diabetes is linked to oral problems, or vice versa. None of our interviewees expressed a deeper understanding of how and why diabetes affects another part of the body, including oral health, except for one person who mentioned, “having diabetes, you don’t heal properly”. Some of them doubted the value of periodontal treatment after having diabetes, in that, “It appeared that he [dentist] tends to want to do gum surgery on every person that walks in the door” (p15) or, some denied the connection between oral health and diabetes, saying, “ I don’t believe that it’s related to my diabetes.” (P22). Reasons behind the low perceived importance of oral care in diabetes included a barrier to access to related information. The most prominent barriers reported were never being told by a healthcare provider and having limited resources to know about the benefits of oral self-care. When asked about whether or not they had heard about the connection between oral health and diabetes from any healthcare providers, including doctors, endocrinologists, nurses, dentists, or dental hygienists, the typical responses were, “Never, never. I have not had one doctor relate diabetes to the condition of my mouth or my teeth (P15)” or, “No, I have not. I don’t think so. they have not yet.”, “I have been a diabetic for years, no one has ever really gave me education on like how to care for my mouth diabetes.” Some reported, “Just my general doctor, telling me that pregnancy can make them (teeth) fall out and deteriorate.” (P26), and “Not generally… when I go to the dentist, they know I’m diabetic. So they always inquire about my A1Cs. And I guess as long as it’s normal or not too high, then you know, they’ve never really said anything.” One inquisitive participant mentioned his failure to find related information online in that, “When I was diagnosed with diabetes, you know, I went to WebMD^®, I went to Mayo Clinic. Those standard types of sources to get quick information.” With further questioning about these internet searches, inquiring, “Did you happen to see anything about oral health?”, the response was, “I did not. No.” (P2). Only one person said, “I have read that … online… someplace, and just off a random article I read when I am online (P15)”.

The last major theme was limited self-efficacy on optimal oral self-care in diabetes. It was very common to hear respondents admitting that they do less than they should or expressing shame, regret, or self-judgment about their ability to perform optimal oral self-care. People reported, “ I don’t do [flossing] as much as I should. I don’t floss nearly like I should. Not at all. (P9)”, “ I would say I don’t take care of my teeth as much as I should.” (P15), “I don’t brush my teeth every day, even though I know I am supposed to.” (P26), “I should be drawn and quartered. I have had dental insurance for years; I haven’t been to the dentist. The dentist will be the lowest on the totem pole. … not having the time and energy. The dentist was on the back burner… I hate the way I look… I mean, my teeth, my mouth.” (P14). In addition, some participants’ lack of self-efficacy in performing oral care was also exhibited in their desire to have additional help from others. For example, one participant said, “I suppose I could enlist the help of my wife to urge me to brush my teeth at night.” (P15). Barriers to self-efficacy in practicing oral self-care were often expressed in the interviews as having confusion about the correct oral health strategy, having confusion about choosing appropriate oral health tools, or having challenges in effectively cleaning between the teeth due to gaps between teeth, making the food get stuck easily. Typical quotes related to this concept are:

“After 10 years, the facts change. You know, like eggs are good for you and eggs are not good for you. That type of thing. So, you know, like, oh, well, electric toothbrushes are good for you. Oh, no. They do much damage. They do damage…So it’s just like OK, just try if we see what… you know, shoot for the best.”

(P14)

“I don’t know whether those [floss picks] are any good compared to flossing yourself with a full piece of floss, you know.”

(P15)

“I’m just not sure if it [Waterpik] is really doing what I needed it to do.”

(P29)

Due to the limited self-efficacy in practicing optimal oral self-care, although some people used multiple oral hygiene tools, instead of having a consistent and systematic oral self-care regimen, they often improvised with oral care and oral hygiene tools. Their oral self-care practices were also largely varied and often combined with multitasking, rushing through, or skipping patterns. Examples of some typical quotes in this regard are:

“It depends. Sometimes I will hold it in my mouth till I maybe get halfway dressed…And then sometimes I spit it out before then. It just depends on what I’m doing. Sometimes it’s longer than others. I would say that routine doesn’t have any set time.”

(P9);

“ In a hurry, less than a minute with the mouthwash. And then, when I really take my time, I take at least 5 min”

(P14);

“try to floss daily, but that doesn’t always happen because life is busy of course.”

(P17);

“[floss or clean between teeth] if I got something stuck in my teeth, like if I was eating corn.”

(P26).

It is noteworthy that even among the participants who seemed to have high perceived self-efficacy on oral self-care, they seem overly confident about their ability to perform optimal oral self-care, as this same group often reported long-standing habits of reflecting less than optimal tooth brushing and flossing. For example, typical quotes from this group were, “Brush twice a day. That’s pretty much it,” (P2); “ I only brush once a day and I don’t floss because my gums [are] in great shape, and my teeth have very little tartar. (P13)”, and “When I do it [brush teeth], I do it right, but I don’t do it [brush teeth] every day. (P12)”.

Three participants, “alert observers”, emphasized that catching early signs can prevent later issues. They described the early signs that had caught their attention as, “anything that’s alarming”, “anything popping or fluid or blood”, “ulcers”, “painful or hurt”. One participant said, “I typically catch things that pop up because I try to be aware of what’s going on, just because I am a diabetic. I don’t want things to get worse if I could catch them.” (P17). Interestingly, these “alert observers” usually did not have high perceived self-efficacy due to higher expectations and the feeling that they were not fully prepared and able to do what is best for them because they had diabetes. They reported, “They tell us how to take care of, you know, our eyes. But like, how is our oral health like? How is that… you know… how does that all make sense?” (P16) and “We know to brush our teeth and things of that nature, but maybe if we’re having symptoms or issues that might be related to diabetes, and you know it’s something that’s not as much talked about.” (P17).

3.4.3. Feedback on DiaOral^© Blueprint

After reviewing the DiaOral^© blueprint on paper, participants would describe the DiaOral^© blueprint as “interesting”, “informational”, “encouraging”, and “beneficial, and helpful”, and describe the DiaOral^© structure as “self-explanatory”, “accessible”, “straightforward”, “user friendly”, and “easy enough”. They reported liking the way that the blueprint on paper indicated that the DiaOral^© website will not have a lot of text and included photos and videos. When asked, “Do you feel like this is something that you are able to use on a computer?”, all participants, even those who said they are not technology savvy, expressed confidence and interest in using it, such as, “ I would not have a problem with that navigating,” “ I do those things online, yeah.” or, “ it looks like a standard website, just like any other”. Several participants mentioned their desire to have a “one click” login, as in “my Google address would get me in, so I would not have to remember another password,” “there is so many websites that I have to have, and it’s just like if one click would get me in access would be one thing, that’s significant.” “Why do I have to establish username and password, the whole login process, it raises questions in my mind?” In addition, very few participants expressed their desire to have some tracking features such as a place for people to write down things, so they can see progress over time. And one participant mentioned that it would be great if it were a mobile app instead of a website.

4. Discussion

Our prior work and studies by other researchers report that disparities in oral health and oral preventive care are prominent among PWD, compared to their counterparts without diabetes in US adults [4,6,30]. Yet these prior quantitative studies do not provide in-depth insights as qualitative studies can do in elucidating underlying factors contributing to the disparities, and how PWD’s beliefs and attitudes might contribute to these disparities. Among the only five prior qualitative studies (including study employed qualitative methods as a part of the design), four were conducted more than ten years ago and only one was among US patients [31,32,33,34,35]. In this qualitative narrative study focusing on PWD in the US, we found that lack of knowledge and low perceived importance of oral self-care related to diabetes self-management is a significant issue among racially diverse groups with diabetes in the US. It resonated with the finding from other countries that most PWD never heard that diabetes and oral health were connected [33,36]. It also suggests that the underlying factors contributing to current issue may be, in part, related to the separation of oral care and diabetes care in the US health system. The physical separation between dentistry and diabetes medical care has hindered the timely and effective delivery of oral self-care education and behavioral intervention to diabetes patients. Many PWD do not have a “dental home” that could provide preventive care regularly and consistently. In one British study, researchers indicated that even when they do obtain preventive care, diabetes patients often feel their dentists are too busy or lack interest in communicating specific oral care issues related to diabetes [34]. Also, patients have expressed their preference to receive this information in primary care settings, especially at the time they receive diabetes care, or when having self-management discussions with nurses or diabetes educators [31,32,34]. However, these older studies were conducted before the advances in digital health technologies, and has focused on how to improve patient-provider communication.

Our study findings reveal the potential of using web-based digital technology to fill the gap. For example, the DiaOral^© website could be used to facilitate the ADA’s emphasis on four efficient diabetes educational moments (at diagnosis, annual checkups, unmet glycemic control goals, and when complications emerge [37]. Most of these educational and behavioral intervention moments take place in primary care. However, primary care providers do not see attending to and advising on dental issues as within their scope of practice. In general, primary care professionals would refer a patient to their dentist when a dental issue emerges. Although previous research has studied a number of approaches to train healthcare providers, especially nurses, in performing basic oral exams and educating patients, these approaches are not readily available in most areas [38,39,40]. In the internet era, this issue potentially could be alleviated using web-based programming to supplement patient education in primary care settings. However, the varied comfort level with searching for and using health-related information online among diverse diabetes populations, especially those socioeconomically disadvantaged, should be considered when developing online oral health educational material.

This study also revealed that another obstacle is that there is a lack of reputable website information that is simple enough in its content to be readily accessible, instead of requiring extensive searching to pull together all the needed information. The idea of developing a website-based oral self-care intervention for PWD answers such a call. The blueprint of DiaOral^© on paper received favorable feedback from participants, suggesting that as long as the design of a website is simple and intuitive enough, diabetes patients could be willing to receive and use an online education and behavioral intervention outside of dental practice. On the other hand, our study findings not only show that people with diabetes have limited self-efficacy in practicing optimal oral self-care as some previous studies found [31,35], but also reveal the psychological mechanism of how people’s perceptions about the importance of oral health to their diabetes control is rationalized in individual oral self-care behaviors and day to day self-care routines. The Health Belief Model is the most widely accepted model that explains how modifiable factors (i.e., knowledge and health literacy) and individual beliefs (i.e., perceived susceptibility and severity of a defined disease and perceived benefit and barrier of performing a health behavior), determines an individual’s health behavior that can lead to improved health outcomes [16,41]. However, in the field of oral health promotion and behavior change, not many theory-guided behavior changes interventions have been implemented and tested [42]. Limited previous studies were either conducted among children and adolescents [18,25,43], or focused on tooth brushing and flossing only [42]. Health behavior interventions require the effort and commitment of participants to (1) identify and abandon old behavioral patterns that are deemed harmful or suboptimal, (2) establish a new behavior pattern that is beneficial and optimal, and (3) further intensify and solidify the new behaviors with repetition over time to form a new habit. Designing and developing an effective web-based oral self-care intervention for PWD, the DiaOral^© program, requires evidence-based implementation of relevant information to focus on the psychological pathway about perceived susceptibility to oral disease related to diabetes and contents and structure designed in a way that goes beyond merely providing health information. A successful behavior intervention program cannot be achieved without a meticulous program design based on scientific theories of behavior change. The use of the HBM helped conceptualize the influence of each health belief construct: barriers, benefits, severity, and susceptibility, on the final behavioral outcome. Our proposed DiaOral^© intervention may be the first approach using the HBM to guide the program design especially for PWD, a population at high risk for oral diseases.

Limitations

This qualitative narrative study design provided a rich and in-depth understanding of the patterns of oral self-care behavior among PWD, and identified the associated factors and patients’ cognitive rationale underpinning their oral health patterns. However, several limitations must be acknowledged.

First, qualitative studies are time consuming and limited by sample size. In this study, we recruited a small group of diverse PWD to inform the early development of DiaOral^© modules. The findings are not intended to obtain statistically generalizable outcome but rather to provide preliminary support for later website prototype development.

Second, in alignment with the principles of RTA, we recognize that our findings are not absolute truths residing in the data, but are actively constructed interpretations stemming from our positionality as seasoned researchers in oral health and diabetes care. While our knowledge facilitated rapport and deeper interpretation, it was also susceptible to a reflexive bias that shaped how we interpret the findings.

Third, while we employed the HBM as a sensitizing framework to ensure theoretical meaningfulness for a future intervention, this deductive lens may have prioritized certain psychological constructs over others.

Finally, although with a sample size of 15 we satisfied the data saturation requirement for a generic qualitative study, given the heterogeneity of the study participants, we may have missed complete data saturation in some subthemes. In addition, all participants reported positive feedback regarding the DiaOral^© blueprint, which may be influenced by social desirability bias. And the perceived usability, accessibility, and ease of use of a website as presented in blueprint cannot be interpreted as firm satisfaction with an actual website. However, we recognize this as a manageable limitation in a non-generalizable qualitative study to inform the early development of DiaOral^© blueprint, ready to be moved forward with website prototype development.

5. Conclusions

This narrative qualitative study provides preliminary support for the theoretical relevance and feasibility of using HBM constructs to guide the understanding and meaningfulness of the PWD’s oral self-care behavior in diabetes management. Major themes about oral self-care among PWD are lack of knowledge about optimal oral care, low perceived importance of oral care in diabetes, and limited self-efficacy related to optimal oral self-care. The DiaOral^© blueprint, as guided by the HBM, was perceived as acceptable and accessible to diverse PWD. Next steps will be to finalize the DiaOral^© website prototype and subsequently test its real-world acceptability, accessibility, and satisfaction among PWD.