Integrating support for tobacco counseling into electronic dental records: A multi-method approach

Abstract

Dentistry has historically seen tobacco usage as a medical problem. As a consequence, dentistry has not adopted or developed effective interventions to deal with tobacco usage. With the expanded use of electronic dental records, we identified an opportunity to incorporate standardized expert support for tobacco cessation counseling during the dental visit. Using results from observations and focus groups of dentists and hygienists, a decision support system was designed that suggested intervention discussion topics. Because dental providers are always pressed for time, our goal was a 3-minute average intervention interval. To fulfill the provider’s need for an easy way to track ongoing interventions, script usage was recorded through a single click. This helped the provider track what he or she had said to the patient about tobacco cessation during previous encounters and to vary the messages. While the individual elements of the design process were not new, the combination of them were very effective in designing a usable and accepted intervention. The heavy involvement of stakeholders in all components of the design gave providers and administrators ownership of the final product, which was ultimately adopted for use in all our dental group’s clinics.

INTRODUCTION

The development of evidence-based knowledge has become a major focus of the dental profession. Consequently, the scale and scope of evidence-based resources to shape dental professionals’ practices (e.g., systematic reviews) have increased. The challenge is how to take the knowledge from the review or guideline and implement it into dental practice. Use of the electronic dental record (EDR) increased from 25% in 2004 to 55.5% in 2007¹. While EDRs have been adopted because they improve tracking and billing for services, they also can provide evidence-based clinical support^{2, 3}.

In this study, we explored the feasibility of embedding an established tobacco cessation intervention in the EDR of HealthPartners Dental Group (HPDG), a large staff-model group practice in Minnesota. Our primary goal was to help practitioners deliver tobacco-cessation services. The setting was particularly conducive to our work because staff dentists have been using an EDR since 2002; thus, electronic record keeping and interaction with the computer in the clinic are well accepted and used for all aspects of patient care.

Tobacco cessation practice guidelines developed by the U.S. Public Health Service, for example, have evolved from a sizeable, growing body of research ^4–6. Regrettably, despite the overwhelming evidence supporting their use, smoking cessation interventions are inconsistently delivered to tobacco users in both medical and dental practices^7–9. There is a great need to improve the integration of tobacco cessation guidelines into practice¹⁰. The emergence of EDRs may offer opportunities to incorporate best practices (i.e., evidence-based dentistry) into clinical practice, going beyond the data capture functionality in the paper chart. It may be possible to use the EDR to help practitioners incorporate best-practice principles into tobacco cessation.

However, to maximize the effectiveness of our EDR-based tobacco cessation intervention, we needed to understand how oral health providers (OHP), who are primarily dentists and hygienists, use the EDR during delivery of care so that we did not negatively affect the existing flow of the patient encounter and frustrate the user. Therefore, before designing and implementing the tobacco cessation intervention in the EDR, we needed to observe and discuss how OHPs use the EDR while providing direct patient care.

METHODS

The dental encounters of interest for this study were new-patient and recall visits in which preventive services, examinations, and medical history updates are performed. HPDG employs roughly 114 dentists and hygienists, who provide both pre-paid and fee-for-service dental and oral care in 16 dental clinics. HPDG provides care for about 100,000 members. HPDG’s EDR is customized and marketed by General Systems Design (Cedar Rapids, Iowa).

In 2008 HPDG saw roughly 7,500 smokers. Both hygienists and dentists were trained before this study (2001) on assessment and follow-up approaches to smoking cessation interventions. However, no support was provided for integration of this training into the dental encounter.

To determine tobacco usage data collection and workflow, we observed a series of dental encounters in different clinics. We used this information to develop focus group discussion points about EDR improvements for supporting dentist- and hygienist-directed tobacco interventions. We felt that the synergy of contextual inquiry^11, which produces observational data unbiased by attitudes, expectations, or opinions, and focus group sessions, that could elucidate observed behaviors would produce the most meaningful insights for designing our intervention. In addition, while the observations captured current tobacco cessation practices, the focus group discussions could help us envision how they could be improved and the EDR should be modified in response. Also, the focus groups gave participating providers ownership of the ideas produced, making adoption more likely after implementation. This information was used to design an automated tobacco expert system to be embedded in the tobacco usage section of the medical history. As a further development tool, a functional mockup of a proposed system was iteratively presented to providers for feedback.

Observations

Following Institutional Review Board approval, a notice was sent to clinic supervisors at each HPDG clinic to announce and explain the EDR project. Supervisors were given 2 weeks to respond to this notice before initiating subject recruitment. Clinic supervisors did not express any opposition.

Recruitment letters were then sent to all 44 dentists and 70 dental hygienists employed in the HPDG clinics. The letters described the study and informed recipients that they might be asked to participate. Recipients were given 2 weeks to email or phone the study coordinator to opt out of being contacted. After the 2-week intermission, the study coordinator called potential participants until 30 practitioners (16 hygienists and 14 dentists) agreed to participate. The hygienists were all women; of the 14 dentists, four were women and 10 were men.

Site selection

Clinic selection was based on sites in which at least one hygienist and one dentist had volunteered for the study. It was important to be able to observe the entire appointment, including the examination. Therefore, hygienists from clinics in which a dentist would not agree to be observed were not recruited, and vice versa. Once a hygienist-dentist team was identified at a given clinic, the coordinator focused on recruiting at another clinic. Nine of the 16 HPDG locations in the greater Minneapolis-St. Paul area were included. This gave us a broad range of contacts with different providers and clinic cultures, which were surprisingly diverse.

Observers

The observers were selected on the basis of their experience as a dental hygienist and dental hygiene educator (PL) or as a systems analyst (WR). The observers’ combined expertise allowed them to capture an accurate representation of both the clinical and technical processes. Before study initiation, both kinds of observers were taught how to use the EDR by the same person who trained HPDG personnel.

Observation process

Twenty-five field observations were carried out by WR and PL at 15 dental hygiene recall and 10 dental hygiene new-patient appointments. Contextual inquiry, an ethnographic method, was used to document work sequence, information acquisition and documentation, work process breakdowns, and verbal feedback ^{12, 13}. Before each observation, the observed clinician asked for the patient’s permission to allow the researchers in the cubicle during the appointment. Once the patient consented, the observers positioned themselves in the cubicle on opposite sides of the patient chair to observe from multiple vantage points. Care was taken to not disrupt the appointment. Numbered notes were made on a sheet with screen shots of the primary EDR screens used in an exam encounter. This allowed a systematic, methodical collection of data about the process of the clinical encounter. Care was also taken to include all screens on which tobacco information could be collected. After the patient left the exam room, the observers asked the clinician about any procedural steps with the EDR that were not clear during the observation. Discussion of tobacco use assessment was not mentioned until the end of all observations in a session.

Field-note processing

On the same day after each observation, PL and WR informed each other, in chronological order, of what each had observed. The observers then synthesized their written observations and elaborated on the details and actions while the information was fresh in their minds. Observer PL returned to the office, where she transcribed the field notes into an electronic document.

Observations were limited to two patient appointments per day. This scheduling facilitated field-note processing within 48 hours of each observation and limited observer fatigue and multiple-observation confusion. The transcribed field notes were subsequently reformulated as formal models, specifically, a flow model and sequence model. In addition, each model type was consolidated across offices and visits. Qualrus Version 2 helped develop themes, which were used as discussion points in the focus groups.

Focus groups

We conducted three focus groups to determine dentists’/hygienists’ attitudes towards the EDR and self-perceived barriers and facilitators to its use to support tobacco counseling. One focus group consisted of hygienists, one contained only dentists, and the third was a mix of both dentists and hygienists. A trained focus group moderator led the focus groups, and discussions were audio taped and transcribed. Two researchers reviewed transcripts, and themes were identified. All participants were de-identified in the transcripts.

Each focus group comprised two activities. First, the attendees were given screen shots of the locations in the EDR where tobacco use was addressed. They were asked to rank the different screens relative to their usefulness and position in the work flow for tobacco cessation counseling. Second, each group discussed current barriers and possible facilitators to improved support for tobacco cessation counseling. Through the two activities, we successfully identified sets of practice patterns that suggested the most effective locations and processes for tobacco clinical reminders in the EDR (Figure 1).

Figure 1.

Focus group participants’ perceptions of most effective locations for tobacco clinical reminders in the EDR, by percentage of total respondents (n=30)

Expert system design

The next step was to use the observations and focus group results to mock up a functioning prototype of an expert system that would fulfill provider needs. We presented this prototype to system administrators, dentists, and hygienists for their feedback. This was an iterative process of presentation and revision between project personnel and stakeholders. An important consideration was our discovery early in the study that we could not modify the EDR locally; rather, we would have to spend considerable money and time to have the system vendor modify it. Therefore, we wanted to get it right the first time.

RESULTS

A total of 36 dentists/hygienists were phoned to recruit enough participants. Six providers declined participation. Reasons given for not wanting to be observed included: 1) Felt that participating might cause them to run behind schedule and increase patient wait times; 2) Believed that the project was really a covert means of “spying” on employees; 3) Hygienist was willing to participate, but dentist was not; 4) Dentist was willing to participate, but hygienist was not; 5) One practitioner reported undergoing chemotherapy; and 6) One practitioner planned to leave HPDG soon.

Assessing practice patterns relative to EDR usage

Twenty-five exam encounters were observed. Observations noted that workflow through EDR screens was fairly consistent. They would start with the health history, proceed to tooth and periodontal assessments, and end with evaluations of future risks. Most providers had favorable attitudes towards EDRs, although they wished the EDR could interact directly with the electronic medical record (EMR) to keep the health history up to date. Hygienists had most of the information acquisition and documentation responsibilities. They updated the health history and evaluated periodontal status. The dentist examined the teeth but addressed their observations to the hygienists, who entered them into the EDR. With input from the hygienists, the dentist would update the caries, periodontal and oral cancer risk assessments.

Barriers to workflow were limited but serious when they occurred. Single-monitor use prevented EDR access while dentists viewed radiographs. During this process, the hygienist was often forced to take paper notes and enter observations after radiograph examination. EDR breakdowns/lockdowns were infrequent but seriously interrupted workflow because providers had to reboot the system.

There were two types of functionality for recording tobacco use in the EDR. The first, in the health history, allowed the hygienist to record information about the patient’s tobacco use habits. This included type of tobacco, frequency of use, amount used, and number of previous quit attempts [Figure 2]. We observed that the only one of these items consistently addressed by the hygienist was whether the patient was still smoking. Because the system carried forward the smoking information from the previous exam, any patient who had taken up smoking or was a smoker recidivist would not be identified or have their status updated. The second place for recording of tobacco use status was in the periodontal risk assessment screen, which was typically performed toward the end of the exam. This was filled out with input from both the dentist and hygienist but was not linked in any programmatic way with the information obtained through the medical history. The second form of functionality related to recording activities centered on counseling or supporting the patient in tobacco cessation. These issues were addressed in the hygiene information form, where hygiene education interventions were recorded, as well as on the treatment plan, where the dentist could refer the patient to tobacco use phone counseling. Identification of barriers to tobacco cessation information use was chosen as an important activity for the focus groups. The other obvious issue to address in the focus groups was, considering workflow, which tobacco-related EDR screen would be the best place for tobacco-intervention suggestions.

Figure 2.

Original medical conditions page of the EDR, with tobacco usage questions highlighted

Our observations documented three main problems with how the EDR supported tobacco use documentation and cessation activities. First, the information related to tobacco use was spread over several separate parts of the EDR that were not logically connected through the clinical workflow. If information was recorded in one place, it was not automatically updated in the others. Second, the answer to current tobacco use on the health history was, by default, carried over to subsequently updated history forms, making the question easy for providers to miss or ignore. In addition, because providers did not understand the purpose of or how to use the follow-up questions for determining level of addiction, they often left these fields blank. Third, if the provider did indeed conduct a tobacco intervention, there was no convenient place in any of the tobacco-related screens to record what was done for future reference. In summary, the EDR poorly supported executing the different steps involved in tobacco use assessment and cessation and made it hard to document them if they were done.

Focus groups for hygienists and dentists on potential tobacco intervention

The hygienists interviewed in the focus group felt that the appropriate place for the collection of tobacco information was the health history. We were also told that they were afraid of offending a patient with no interest in quitting, and that this was more likely when they did not know the patient’s previous responses to cessation approaches. Therefore, they felt strongly that they needed some mechanism at this location to record notes about past reactions to questions about tobacco cessation. Participants felt that, while it was up to them to collect the tobacco status information, the dentist should be involved in the cessation activities.

A very interesting finding was that hygienists felt that patients should be referred to the state-sponsored, telephone-based tobacco quitting support resources more often than it was being done. However, we heard as many excuses as there were participants for why they would be unable to get the form from the printer or access the fax machine to do so. We concluded that this important process needs to be simplified and streamlined.

The dentists’ perspective on tobacco interventions was different from that of the hygienists. Dentists’ suggestions tended to relate to using the tobacco information collection fields available in the health history. However, because these fields were completed by the hygienist, the dentists did not seem to know what was in them. They suggested that tobacco interventions be chosen by the dentist at or near the end of the encounter and printed together with other selected interventions. The dentists’ perspective was probably related to their interaction with the patient in the flow of the encounter. The dentist examined the patient after the medical history was taken. Therefore, if the hygienist did not point out the patient’s smoking status, the dentist might be unaware of it.

Design of changes to the EDR to support a tobacco intervention

Using the information collected in the observations and focus groups, we modified the existing system (Figure 3) that would provide much of the missing support to providers. This involved creating an expert support system in the health history to evaluate the smoker’s tobacco use and suggest approaches for intervention and tracking. To do this using the tobacco use information collected, we automated the calculation of two of the most accepted approaches to classifying patients’ positions relative to presenting tobacco cessation messages: desire to quit and level of addiction.

Figure 3.

Modified medical conditions page of the EDR, with tobacco usage questions highlighted

Dental providers are always under time constraints, and the addition of a tobacco intervention for the roughly 15% of their patients who smoke was clearly a factor in limiting provision of tobacco support. Therefore, the automation process was intended to encourage a tobacco intervention while providing the necessary support for a quick and accurate process. Our goal was a 3-minute average intervention interval. To this end, we gave the providers scripts to use with the patients (Figure 4) based on the desire to quit, tobacco usage, and level of addiction. To fulfill the provider’s need for an easy way to track ongoing interventions, script usage was recorded through a single click (See example process flow; Figure 4). This helped the provider track what he or she had said to the patient about tobacco cessation during previous encounters and to vary their messages.

Figure 4.

Tracking system and scripts provided in modified EDR

DISCUSSION

Dentistry has historically seen tobacco usage as a medical problem⁹. As a consequence, dentistry has not adopted or developed effective interventions to deal with tobacco usage. With the expanded use of EDRs, we have an opportunity to incorporate standardized expert support for tobacco cessation counseling during the dental visit. This paper addresses an approach to developing an intervention in which providers help shape a tobacco cessation program. By involving providers, we were able to incorporate their insights and increase their acceptance of an EDR-based tobacco cessation intervention.

The methods we used to design an EDR-based dental tobacco cessation strategy were not new, but the way we used them was. We did not take the typical approach that we, as experts and developers, knew the best approach to design and implement our system. Rather, we first learned by observing the current practices of dentists and hygienists as they interacted with patients on tobacco cessation in the EDR. We analyzed these observations using contextual inquiry and qualitative analysis methods. Subsequently, we used the results as the input for focus group discussions grounded in actual practice as we observed it. This allowed the participants to critically reflect on what clinicians were currently doing, why they were doing it, and how the EDR could support a more ideal practice. By taking this approach, we successfully combined the expert knowledge of researchers with the experience and needs of practitioners. We made practitioners partners in the process, and they became stakeholders in its success.

Our main finding in this study was the degree and manner in which the existing EDR impeded clinical work on tobacco cessation. The participating practitioners did not argue against the appropriateness of providing tobacco cessation in dental practice, nor did they have any problem with doing so themselves. Aside from time pressures, most barriers to tobacco cessation interventions identified by our study were logistical. Of note, most logistical barriers were due to the design of the EDR.

As discussed, the EDR made effective tobacco documentation and intervention difficult in a number of ways. It did not help the provider learn about tobacco use status, help systematically perform tobacco status assessment and cessation interventions, or support effective and efficient documentation and information review. However, through observation, provider focus groups, and an iterative provider-tested design process, we were able to design an effective tobacco intervention support process. The most satisfying outcome of this approach is that HPDG has adopted the resulting tobacco intervention support process in all its clinics.