An Intuitive Risk Communication Tool to Enhance Patient–Provider Partnership in Diabetes Consultation

Abstract

Introduction:

This technology report introduces an innovative risk communication tool developed to support providers in communicating diabetes-related risks more intuitively to people with type 2 diabetes mellitus (T2DM).

Methods:

The development process involved three main steps: (1) selecting the content and format of the risk message; (2) developing a digital interface; and (3) assessing the usability and usefulness of the tool with clinicians through validated questionnaires.

Results:

The tool calculates personalized risk information based on a validated simulation model (United Kingdom Prospective Diabetes Study Outcomes Model 2) and delivers it using more intuitive risk formats, such as “effective heart age” to convey cardiovascular risks. Clinicians reported high scores for the usability and usefulness of the tool, making its adoption in routine care promising.

Conclusions:

Despite increased use of risk calculators in clinical care, this is the first time that such a tool has been developed in the diabetes area. Further studies are needed to confirm the benefits of using this tool on behavioral and health outcomes in T2DM populations.

Keywords

primary care risk communication risk perceptions shared decision making type 2 diabetes mellitus

Good patient–provider communication and shared decision-making have been associated with improved self-management^1,2 and glycemic control³ in type 2 diabetes mellitus (T2DM) populations. Although significant progress has been made to enhance this “patient–provider partnership” over the past two decades,⁴ some challenges remain. For example, many patients still face difficulties in understanding the probabilistic risk information traditionally used by clinicians to convey increased risks of complications associated with poor glycemic control.⁵ Given that risk perceptions play a key role in the behavior change process,^6,7 and that people with T2DM tend to have inaccurate perceptions of these risks⁵, there is a need to develop risk communication strategies that are more effective than existing methods in correcting for risk misperceptions.^8,9 Most interventions to date have hardly produced any effect on risk perception accuracy, with the most effective strategies only leading to short-term improvements lasting between two⁸ and six weeks.¹⁰ This article reports on the development of a digital tool designed to support providers in communicating diabetes-related risks more intuitively in routine care, so that patients can update their risk perceptions and, in turn, engage in shared decision making more effectively.

Methods

A team including two digital health specialists, two health economists, one behavioral economist, and one general practitioner experienced with diabetes research (“the research team”) took part in the development process. This consisted of three steps: (1) selecting the content of the tool; (2) developing the digital interface; and (3) testing the usability of the interface with a sample of target end users (ie, providers) and refining it accordingly.

Content of the Tool

Selection of the content and format of the risk message (what should be communicated, and how) was guided by a thorough review of the risk communication literature and by insights from behavioral science. Evidence suggests that the risk message should stay brief and simple (so that it can be understood regardless of patient numeracy skills),¹¹ and should be personalized.¹² In addition, the use of imagery and interactive features, which stimulate attention and working memory, has been recommended.^13,14

Based on these recommendations and on the results of a feasibility study conducted in routine care with 40 patients with T2DM, in which specific innovative risk metrics were found to be superior over probabilistic information in increasing risk perception accuracy,¹⁵ we chose to limit the content of the tool to just two types of risk information, conveyed through carefully selected formats.

Increased cardiovascular risks, expressed through “effective heart age”

Despite being the major threat associated with lack of diabetes control, increased cardiovascular (CV) risks tend to be largely underestimated by people with T2DM.⁵ We chose to use the concept of effective heart age to express those risks, which corresponds to the age of a similar “well-controlled” individual (ie, whose risk factors are within normal value ranges) with matching CV risk probability.¹⁶ This format makes the risk information more salient and memorable as compared to communication of (seemingly abstract) probabilistic information.^17,18

Reduced life expectancy, conveyed using a short time frame

Communicating reductions in life expectancy associated with lack of diabetes control has the advantage of capturing a more global impact of the detrimental effects on health (ie, conveying the “big picture”). We chose to use a shorter (monthly) time frame to communicate such losses of lifetime, because positive effects on risk understanding¹⁹ and perceptions¹⁰, ²⁰ have been reported. For example, a 0.5% (14 mg/dL) increase in HbA1c may translate into a 3-day loss of lifetime each month, rather than a 2-year loss at the end of life. This may be interpreted as “aging faster” than a similar, well-controlled individual.²¹ The superiority of shorter time frames, which make losses look more immediate, may be explained by recent findings in behavioral science where many people with T2DM were found to express a present bias (the tendency to value immediate rewards more than future, bigger rewards)²² and loss aversion (the tendency to dislike losses more than enjoying equivalent gains).²³

Digital Interface

We then developed a digital interface based on the selected content described in the previous section. The code uses simulation data produced by the United Kingdom Prospective Diabetes Study (UKPDS) Outcomes Model (version 2) to calculate personalized risk estimates.²⁴ To maximize the user-friendliness of the interface, we collaborated with a digital studio specialized in user experience design and with two volunteer patients. Two design criteria were considered of higher priority: ease of use and speed of use, in order to minimize the risk of providers being reluctant to adopt the tool in routine care.

Usability Testing

Finally, we tested the usability of the interface with a sample of providers. Following Li et al (2013),²⁵ our protocol consisted of a three-step iterative process whereby we (1) tested the pilot version to identify modifiable usability issues (Cycle 1); (2) stopped testing after no new issue could be identified and made revisions; and (3) tested the revised version to identify new issues (Cycle 2). A usability issue was identified when a participant was unable to move on to the next step due to design or programming errors, or when at least two participants commented on the same issue.

Participants were given a virtual patient profile and asked to use the tool in a simulated way while expressing their thoughts aloud. Notes were taken and sessions were audio-recorded. The participants were also asked to complete two questionnaires: (1) the System Usability Scale (SUS)²⁶ for ease of use; and (2) a tailored questionnaire rooted in the Technology Acceptance Model (TAM) for intentions to use the tool, if made available.²⁷ We used descriptive statistics to summarize participant characteristics and SUS- and TAM-based scores after each testing cycle. No statistical comparisons were conducted between cycles, as hypothesis testing was not a goal of this study. Our data analysis was inductive, as we sought to understand the participants’ experience with the tool rather than to test a preconceived theory.

Results

The tool was given the name “PERCODIA”, which stands for “personalized risk communication for people with type 2 diabetes”. It is available in English language and compatible with both computers and electronic tablets. The digital interface is made up of four pages: a welcome page, a data-entry page, and two pages displaying heart age and speed of aging. It is intended to be displayed on a screen that is viewed simultaneously by both the clinician and the patient. Figures 1 and 2 illustrate the data-entry and heart age pages, respectively; other pages and a more detailed description can be found in Supplementary Materials. In addition, Table 1 provides the list of information needed to calculate the patients’ personalized risk estimates, and Table 2 indicates the modifiable risk factors that can be targeted by clinicians to visually show patients the potential improvement (or degradation) in risk estimates associated with an improvement (or degradation) of those risk factors.

Figure 1.

Data-entry page.

Figure 2.

Heart age page.

Table 1.

Information Needed by Tool Users to Calculate Patients’ Personalized Risk Estimates.

Type	Unit	Range
Sex	Categorical	“Male” or “Female”
Duration of T2DM	Year	<5; 5-10; >10
Age	Year	35-75
Height	Centimeter	140-210
Weight	Kilogram	40-140
Smoking status	Categorical	“Yes” or “No”
HbA1c	%	6.5-12.5
	mmol/mol	47-113
Systolic blood pressure	mm Hg	130-190
LDL-cholesterol	mmol/L	2.6-5.0
HDL-cholesterol	mmol/L	0.0-2.0
Estimated glomerular filtration rate	mL/min/1.73 m²	<60; ≥60

Abbreviations: HbA1c, glycated hemoglobin; HDL, high-density lipoprotein; LDL, low-density lipoprotein; T2DM, type 2 diabetes mellitus.

Table 2.

Modifiable Risk Factors that can be Targeted by Tool Users to Illustrate the Potential Improvement (or Degradation) in Patients’ Personalized Risk Estimates.

Risk factor	Variation	Range
Smoking status	Qualitative (box to tick)	From “Yes” to “No”; from “No” to “Yes”
HbA1c	Quantitative (slider)	6.5-12.5 (%)
LDL-cholesterol	Quantitative (slider)	2.6-5.0 (mmol/L)
Weight	Quantitative (slider)	40-140 (Kilogram)

Abbreviations: HbA1c, glycated hemoglobin; LDL, low-density lipoprotein.

Usability Testing

We recruited 16 providers on a voluntary basis in Oxford, UK. Twelve were general practitioners and four were junior medical practitioners engaged in postgraduate training. Of those, 11 participated in Cycle 1 and 5 tested the revised version in Cycle 2. Seven modifiable usability issues were identified in Cycle 1 (see details in Supplementary Materials), whereas Cycle 2 did not reveal any new issue. All SUS- and TAM-based scores are reported in Table 3.

Table 3.

Usability Testing Results: SUS- and TAM-based Scores.

	Cycle 1	Cycle 2	Cronbach’s α^a
Participants, n (%)
General practitioner	9 (82)	3 (60)
Junior doctor	2 (18)	2 (40)
SUS items, mean scores [SD] ^b
1. I think that I would use this system frequently	4.3 [0.8]	4.4 [0.5]
2. I found the system unnecessarily complex	1.5 [1.2]	1.2 [0.4]
3. I thought the system was easy to use	4.7 [0.5]	5.0 [0.0]
4. I think that I would need the support of a technical person to be able to use this system	1.1 [0.3]	1.2 [0.4]
5. I found the various functions in this system very well integrated	3.8 [1.3]	4.6 [0.5]
6. I thought there was too much inconsistency in this system	1.2 [0.4]	1.0 [0.0]
7. I would imagine that most people would learn to use this system very quickly	4.9 [0.3]	4.6 [0.5]
8. I found the system very cumbersome to use	1.3 [0.5]	1.0 [0.0]
9. I felt very confident using the system	4.3 [0.9]	4.6 [0.5]
10. I needed to learn a lot of things before I could get going with this system	1.2 [0.4]	1.0 [0.0]
SUS score, mean [SD] ^c	89.3 [4.5]	94.5 [4.3]
TAM-based items, mean scores [SD] ^b
Perceived usefulness			0.83
1. Using PERCODIA can enable me to complete my patient care more quickly	3.0 [1.1]	3.2 [0.8]
2. Using PERCODIA can improve my patient care and management ^d	4.5 [0.7]	4.4 [1.3]
3. Using PERCODIA can make my patient care and management easier	4.2 [1.0]	4.4 [0.5]
4. I would find PERCODIA useful for my patient care and management^d	4.5 [0.5]	5.0 [0.0]
Total	4.0 [0.7]	4.3 [0.5]
Attitude			0.84
1. Using PERCODIA in patient care and management is a good idea	4.3 [0.6]	4.8 [0.4]
2. Using PERCODIA in patient care and management is pleasant^d	4.6 [0.7]	4.8 [0.4]
3. Using PERCODIA is beneficial to my patient care and management	4.3 [0.8]	4.6 [0.5]
Total	4.4 [0.6]	4.7 [0.2]
Intention to use			0.82
1. I intend to use PERCODIA in my patient care and management when it becomes available in my surgery	4.1 [0.8]	4.8 [0.4]
2. I intend to use PERCODIA in my patient care and management routinely^d	4.4 [0.8]	5.0 [0.0]
Total	4.2 [0.7]	4.9 [0.2]

Calculated based on Cycle 1 scores; ^bStrongly disagree = 1; Strongly agree = 5. SUS scores were calculated by summing the score contribution for each item and multiplying the sum of these score contributions by 2.5. Total SUS scores range from 0 to 100, with a higher score indicating a higher degree of user-friendliness; ^cFor terms 1, 3, 5, 7, 9, the score contribution is the scale position minus 1; for items 2, 4, 6, 8, 10 it is 5 minus the scale position; ^dReverse scored items (these items were negatively phrased to increase respondents’ attention).

Abbreviations: PERCODIA, personalized risk communication for people with type 2 diabetes; SD, standard deviation; SUS, system usability scale; TAM, technology acceptance model.

Discussion

This research aimed to offer clinicians a practical tool to communicate diabetes-related risks more intuitively in routine care, so that shared decision making is made easier. Quoting Greenhalgh and colleagues, clinical tools that contain quantitative elements must be designed “to support conversations, not climb probability trees”, because real shared decision making involves introducing evidence “in a way that informs a dialogue about what best to do, how, and why”.²⁸ With PERCODIA, providers have the option to visually show patients how reducing a modifiable risk factor can improve heart age or slow down speed of aging, and how each of these two metrics can be used as an intuitive marker to set risk reduction goals.

Providers who tested the tool reported high scores for usability and usefulness, making its adoption in routine care promising. Nevertheless, the development process had a few limitations. There was a risk of selection bias due to the recruitment of participants on a voluntary basis, and budgetary constraints prevented us from integrating more sophisticated features into the interface. In practice, a function to auto-populate patient data through linkage to electronic health records would be beneficial to save additional time in consultation. Finally, although the tool calculates risks based on the estimates of a successfully validated model,²⁹ its own validation as a predictive model is still needed.

Provided that future studies investigating the effects of using PERCODIA show benefits on risk perception accuracy and shared decision making, there will be an opportunity to improve the management of people with T2DM by adding this tool to the clinician’s toolbox in routine care. Such risk assessment tools have shown promise in improving behavioral outcomes or risk factor levels in other disease areas,³⁰ and several clinical guidelines now recommend their use as part of the routine care pathway (eg, prevention of hypertension³¹ or atherosclerotic disease³²). By contributing to the development of such tools in the diabetes area, we hope to support clinicians in helping patients with T2DM achieve significant risk factor changes and, eventually, improvements in health outcomes.

Conclusions

Despite increased use of risk calculators in clinical care, this is the first time that such a tool has been developed in the diabetes area. Future work will explore the effects of using this tool on patient and provider behavioral outcomes in routine care. If proven beneficial, positive effects on health outcomes in T2DM populations are to be expected.

Supplemental Material

sj-pdf-1-dst-10.1177_1932296821995800 – Supplemental material for An Intuitive Risk Communication Tool to Enhance Patient–Provider Partnership in Diabetes Consultation

Supplemental material, sj-pdf-1-dst-10.1177_1932296821995800 for An Intuitive Risk Communication Tool to Enhance Patient–Provider Partnership in Diabetes Consultation by Thomas Rouyard, José Leal, Dario Salvi, Richard Baskerville, Carmelo Velardo and Alastair Gray in Journal of Diabetes Science and Technology

Footnotes

Acknowledgements

The authors would like to thank Dr. James Groves at Oxford University Innovation Ltd, Mariana Morris at Fruto UX Design Studio, and all the volunteers who kindly participated in the development process.

Authors’ Note

Data included in this manuscript were presented in part at the 2020 BEST Conference on Human Behaviour & Decision Making (Brisbane, Australia) and at the 2020 CBC Online Conference “Behaviour Change for Health: New and Emerging Science and Technologies” (London, UK).

Abbreviations

CV, cardiovascular; SUS, system usability scale; T2DM, type 2 diabetes mellitus; TAM, technology acceptance model; UKPDS, United Kingdom Prospective Diabetes Study.

Declaration of Conflicting Interests

The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: the working title of intellectual property (IP) associated with the technology presented in this report falls within a sharing agreement between the authors and Oxford University Innovation Ltd. Oxford University Innovation Ltd is a wholly owned subsidiary of the University of Oxford in charge of managing the University’s IP portfolio.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Financial support for this research was provided entirely by grants from the European Institute of Innovation and Technology (EIT Health) and the UK National Institute for Health Research (NIHR), Collaboration for Leadership in Applied Health Research and Care (CLAHRC) at Oxford Health NHS Foundation Trust. The funding agreement ensured the authors’ independence in designing the study, interpreting the data, writing, and publishing the report.

ORCID iD

Thomas Rouyard

Supplemental Material

Supplemental material for this article is available online.

References

Ratanawongsa

Karter

Parker

, et al. Communication and medication refill adherence: the Diabetes Study of Northern California. JAMA Intern Med. 2013;173:210-218.

Heisler

Cole

Weir

, et al. Does physician communication influence older patients’ diabetes self-management and glycemic control? Results from the Health and Retirement Study (HRS). J Gerontol A Biol Sci Med Sci. 2007;62:1435-1442.

Aikens

Bingham

Piette

JD.

Patient–provider communication and self-care behavior among type 2 diabetes patients. Diabetes Educ. 2005;31:681-690.

Tamhane

Rodriguez-Gutierrez

Hargraves

Montori

VM.

Shared decision-making in diabetes care. Curr Diabetes Rep. 2015;15(12):112.

Rouyard

Kent

Baskerville

, et al. Perceptions of risks for diabetes-related complications in Type 2 diabetes populations: a systematic review. Diabet Med. 2017;34:467-477.

Conner

Sparks

. Theory of planned behaviour and health behaviour. In: Conner

Norman

, eds. Predicting Health Behaviour. Buckingham: Open University Press; 2005:171-222.

Sheeran

Harris

Epton

Does heightening risk appraisals change people’s intentions and behavior? A meta-analysis of experimental studies. Psychol Bull. 2014;140:511.

Welschen

Bot

Kostense

, et al. Effects of cardiovascular disease risk communication for patients with type 2 diabetes on risk perception in a randomized controlled trial: the @RISK study. Diabetes Care. 2012;35:2485-2492.

French

Cameron

Benton

Deaton

Harvie

Can communicating personalised disease risk promote healthy behaviour change? A systematic review of systematic reviews. Ann Behav Med. 2017;51(5):718-729.

10.

Asimakopoulou

Fox

Spimpolo

Marsh

Skinner

TC.

The impact of different time frames of risk communication on type 2 diabetes patients’ understanding and memory for risk of coronary heart disease and stroke. Diabet Med. 2008;25(7):811-817.

11.

Fagerlin

Zikmund-Fisher

Ubel

PA.

Helping patients decide: ten steps to better risk communication. J Natl Cancer Inst. 2011;103:1436-1443.

12.

Ahmed

Naik

Willoughby

, et al. Communicating risk. BMJ. 2012;344:e3996.

13.

Rothman

Kiviniemi

MT.

Treating people with information: an analysis and review of approaches to communicating health risk information. J Natl Cancer Inst Monogr. 1999;1999:44-51.

14.

Spiegelhalter

Pearson

Short

Visualizing uncertainty about the future. Science. 2011;333:1393-1400.

15.

Rouyard

Leal

Baskerville

, et al. Nudging people with type 2 diabetes towards better self-management through personalized risk communication: a pilot randomized controlled trial in primary care. Endocrinol Diab Metab. 2018;1:e00022.

16.

Wells

Kerr

Eadie

, et al. “Your Heart Forecast”: a new approach for describing and communicating cardiovascular risk? Heart. 2010;96:708-713.

17.

Soureti

Hurling

Murray

, et al. Evaluation of a cardiovascular disease risk assessment tool for the promotion of healthier lifestyles. Eur J Cardiovasc Prev Rehabil. 2010;17:519-523.

18.

Bonner

Bell

Jansen

, et al. Should heart age calculators be used alongside absolute cardiovascular disease risk assessment? BMC Cardiovasc Disord. 2018;18:1-8.

19.

Heard

Rakow

Spiegelhalter

Comparing comprehension and perception for alternative speed-of-ageing and standard hazard ratio formats. Appl Cogn Psychol. 2018;32:81-93.

20.

Galesic

Garcia-Retamero

Communicating consequences of risky behaviors: life expectancy versus risk of disease. Patient Educ Couns. 2011;82:30-35.

21.

Spiegelhalter

Using speed of ageing and “microlives” to communicate the effects of lifetime habits and environment. BMJ. 2012;345:e8223.

22.

Mørkbak

Gyrd-Hansen

Kjær

Can present biasedness explain early onset of diabetes and subsequent disease progression? Exploring causal inference by linking survey and register data. Soc Sci Med. 2017;186:34-42.

23.

Rouyard

Attema

Baskerville

, et al. Risk attitudes of people with ‘manageable’ chronic disease: an analysis under prospect theory. Soc Sci Med. 2018;214:144-153.

24.

Hayes

Leal

Gray

, et al. UKPDS outcomes model 2: a new version of a model to simulate lifetime health outcomes of patients with type 2 diabetes mellitus using data from the 30 year United Kingdom Prospective Diabetes Study: UKPDS 82. Diabetologia. 2013;56:1925-1933.

25.

Adam

Townsend

, et al. Usability testing of ANSWER: a web-based methotrexate decision aid for patients with rheumatoid arthritis. BMC Med Inform Decis Mak. 2013;13:131.

26.

Brooke

. SUS: a quick and dirty usability scale. In: Jordan

Thomas

McClelland

Weerdmeester

, eds. Usability Evaluation in Industry. London, UK: Taylor & Francis; 1996:189-194.

27.

Chau

Sheng

ORL

, et al. Examining the technology acceptance model using physician acceptance of telemedicine technology. J Manag Inf Syst. 1999;16:91-112.

28.

Greenhalgh

Howick

Maskrey

Evidence based medicine: a movement in crisis?

BMJ. 2014;348:g3725.

29.

Leal

Hayes

Gray

, et al. Temporal validation of the UKPDS outcomes model using 10-year posttrial monitoring data. Diabetes Care. 2013;36:1541-1546.

30.

Karmali

Persell

Perel

, et al. Risk scoring for the primary prevention of cardiovascular disease. Cochrane Database Syst Rev. 2017;3.

31.

Whelton

Carey

Aronow

, et al. 2017 ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/ASH/ASPC/NMA/PCNA guideline for the prevention, detection, evaluation, and management of high blood pressure in adults: a report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. J Am Coll Cardiol. 2018;71(19):e127-e248.

32.

Grundy

Stone

Bailey

, et al. 2018 AHA/ACC/AACVPR/AAPA/ABC/ACPM/ADA/AGS/APhA/ASPC/NLA/PCNA guideline on the management of blood cholesterol: executive summary: a report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. J Am Coll Cardiol. 2019;73(24):3168-3209.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.72 MB