Cost-Effective Quality of Life Improvement While Reducing Health Care Professional Burnout With an AI-Driven Intervention for Personalized Medicine

Abstract

Background and Aims:

Burnout affects >50% of physicians and nurses. Spotlight-AQ is a personalized digital health platform designed to improve routine diabetes visits. We assessed cost-effectiveness, visit length, and association with health care professional (HCP) burnout.

Materials and Methods:

Complete case within-trial cost-effectiveness analysis embedded within a multicenter, parallel-group randomized controlled trial. Adults with diabetes were recruited from primary and secondary care. Intervention group participants completed the Spotlight-AQ pre-clinic assessment before each routine visit. Health status was assessed with EQ-5D-5L to calculate quality-adjusted life years (QALYs). Client Service Receipt Inventory measured downstream resource use. Total costs and QALYs were calculated using baseline-controlled seemingly unrelated regression with bootstrapping. Haemoglobin (HbA_1c) data were collected. Health care professionals completed the Maslach Burnout Inventory at baseline and study end.

Results:

A total of 98 adults (49 intervention) and 18 HCPs participated. Total costs: £243 (US$310) intervention arm versus £230 (US$293) control arm; incremental cost: £13 (US$16). Total QALYs: 0.362 intervention arm and 0.358 control arm, with an incremental QALY: 0.004. Spotlight-AQ intervention dominated usual care with a 68% probability of cost-effectiveness at a threshold of £30 000 (US$38 294) per QALY gained. Health care professionals reported reduced burnout, emotional exhaustion, depersonalization, and a greater sense of personal achievement. Doctors are more so than nurses.

Conclusion:

Spotlight-AQ has demonstrated cost-effective while delivering improved care and reduced HCP burnout.

Trial Registration:

ISRCTN15511689, registration date: November 1, 2021.

Introduction

Personalized health care is crucial to effective diabetes management and support. To deliver personalized health care, it is vital that health care professionals (HCPs) are adequately supported to address the concerns that matter most to their patients. These go beyond the physical aspects of health and its biomarkers, such as HbA_1c, time in target glycemic range, blood pressure and lipids. The psychosocial aspects of diabetes, including quality of life, social relationships, and competing priorities, must be incorporated into routine care; however, this can be challenging in health care systems based on the medical model of care delivery. As a result, HCP burnout has reached record levels, with over half of physicians and nurses experiencing intolerable work-related stress.^1,2

Recognizing and incorporating the unique needs of each person with diabetes and simultaneously providing support that facilitates effective self-management behaviors for optimal physical, mental, and social well-being is challenging and requires a fundamental shift by HCPs and health systems. With 1.2 million new diagnoses of diabetes every year in the United States alone, ³ a move toward a biopsychosocial model encompassing physical, mental, and social well-being is necessary to avoid overload of the system and a return to affordable, high-quality health care. ⁴

International Diabetes Federation data show that 10% of current global health expenditure is spent on diabetes (US$760 billion), predicted to rise to $825 billion by 2030. ⁵ Indirect diabetes-related and societal costs from premature death, disability, and other health complications add approximately 35% to global health expenditure each year. ⁵ There are additional, often intangible, and less quantifiable personal costs including worry, anxiety, discomfort, pain, loss of independence, concerns about managing diabetes, and fears for future complications and their potential impact on quality of life. Depression, for example, is two to three times more prevalent in people with diabetes than the general population. ⁶

Burnout among HCPs is a key challenge affecting health care practice, safety, and quality, currently influencing 20% to 30% of front-line health care workers to leave medicine. ⁷ Approximately half of US doctors experience substantial symptoms of burnout, with burnout almost twice as prevalent among US doctors than among workers in other fields. ⁸ Nurses also experience a similarly high prevalence of burnout and depression, with 43% reporting high degrees of emotional exhaustion. There are significant correlations between a doctor’s sense of depersonalization and patient satisfaction with their hospital care, and between a doctor’s job satisfaction and patient satisfaction with their health care and patient-reported adherence to medical advice. ⁸

The aim of the current study was to evaluate the impact on HCP burnout and assess the cost-effectiveness of the Spotlight-AQ clinical tool, which consists of a pre-clinic assessment identifying patient priority concerns and mapping those to evidence-based resources to address those concerns and improve routine visits.

Methods

Study Design

We conducted an exploratory multicenter, parallel-group, randomized controlled trial in primary care (GP) and secondary care (endocrinology and diabetes) NHS sites to assess the clinical effectiveness of Spotlight-AQ in adults with type 1 diabetes or type 2 diabetes attending routine outpatient appointments for their diabetes care. The results have been published previously ⁹ but in brief, following provision of informed consent, participants were randomized on a 1:1 basis using computerized randomization software. The Spotlight-AQ intervention is a clinical tool designed to improve routine visits for health care providers and people with diabetes. It consists of pre-clinic questionnaires and mapped resources to address patient-identified holistic priority concerns. Those randomized to the intervention group were asked to complete study questionnaires every three months and the Spotlight-AQ pre-clinic assessment (approximately 3.5 minutes to complete) within a week prior to their scheduled in-person or remote routine outpatient diabetes appointment. The results were discussed during the outpatient visit along with mapped care pathways in a partnered best-fit action plan.

Assessment of Health Care Provider Burnout

Health care professionals completed the Maslach burnout inventory at baseline and at six-month follow-up. This validated questionnaire consists of 22 items across emotional exhaustion, depersonalization, burnout, and sense of personal achievement. ¹⁰ Visit length was calculated for all participants.

Health Economic Evaluation

A within-trial economic evaluation was conducted following the reference case requirements by the National Institute for Health and Care Excellence (NICE). ¹¹ A cost-utility analysis reported health outcomes as quality-adjusted life years (QALYs). A cost-effectiveness analysis reported health outcomes in natural units (absolute mmol/mol and percentage change in HbA_1c). Costs were measured from the perspective of the health care system (price year: 2021/22). The follow-up duration was six months. Future costs and health outcomes were not discounted.

For the cost-utility analysis, health status was measured by the EQ-5D-5L at baseline and six months. EQ-5D-5L responses, age, and gender were used to estimate health utility values at baseline and follow-up following the recommended approach by Hernández Alava et al^12,13 QALYs per participant were estimated by the area under the curve method as per the equation:

$$QALY=(Utilit{y}_{baseline}+Utilit{y}_{6months})\times 0.5$$

For the cost-effectiveness analysis, health outcomes were measured as the change in HbA_1c over six months. The absolute difference in HbA1c was calculated by subtracting HbA1c at baseline from HbA1c at six months. The percentage difference in HbA1c was calculated as the percentage change in HbA1c from baseline to six months. Resource use was estimated retrospectively at baseline and follow-up using the Client Service Receipt Inventory (CSRI) over a three-month recall window. Relevant unit costs were assigned to the quantities of resource consumed. The unit cost of the Spotlight-AQ intervention was assumed to be £10 (US$12) per user. The duration of each consultation was observed in the trial. The observed consultation duration was compared with the expected consultation duration to determine the resource impact on routine consultation length. Table 1 reports the unit costs for all resource items.

Table 1.

Unit Cost of Resource Items.

Item	Unit cost	Description	Source
GP Visit	£41	GP consultation lasting 9.22 minutes	PSSRU 8
Practice Nurse Visit	£8.67	10-minute consultation with a practice nurse at £52 per hour.	PSSRU 8
District Nurse Visit	£31.50	30-minute consultation assuming a Band 6 nurse at £63 per hour.	PSSRU 8
Social Worker Visit	£25	30-minute consultation with an adult social worker at £50 per hour	PSSRU 8
Psychologist Visit	£229	Non-consultant led clinical psychologist (Service Code: 656)	National Schedule of NHS Costs 9
Physiotherapist Visit	£144	One-to-one physiotherapist consultation	PSSRU 8
Occupational Therapist Visit	£118	One-to-one occupational therapist consultation	PSSRU 8
Dietician Visit	£100	One-to-one dietician consultation	PSSRU 8
Specialist Nurse Visit	£29.50	Cost of a 30-minute consultation with a Band 6 hospital-based nurse at £59 per hour	PSSRU 8
A&E Visit	£242	Weighted average of unit costs for A&E currency codes excluding dental and dead on arrival	National Schedule of NHS Costs 9
Inpatient Visit	£378	Most recent excess bed day cost (2017/18) inflated to 2021/22	National Schedule of NHS Costs 9
Outpatient Visit	£235	Average unit cost of one outpatient visit	PSSRU 8
Hospital Day Visit	£1040	Weighted average of unit cost for day case procedures in 2021/22	National Schedule of NHS Costs 9
Spotlight-AQ Intervention	£10	Average cost per patient of using the intervention	Assumption
Diabetes appointment	£6.68	Cost per minute of a diabetes outpatient appointment	Grant (2015) 10

Seemingly unrelated regression was used to estimate the mean cost and mean health outcomes (QALYs per participant and change in HbA_1c per participant) for the intervention and control arms. ¹³ The dependent variables were: total QALYs per participant, change in HbA_1c per participant, and total cost at follow-up. The independent variables were: treatment allocation, baseline health utility, baseline total cost, age, gender, type of diabetes, and baseline HbA_1c. The incremental cost and incremental health outcomes were calculated as the difference in mean costs and QALYs or change in HbA_1c between the arms. For the cost-utility analysis, the ratio of incremental cost to incremental QALYs was compared with the threshold used by NICE to determine cost-effectiveness (£30 000 (US$38 294) per QALY gained).

Multiple imputation by chained equations was used to handle missing data. ¹⁴ The number of imputed datasets (n = 45) corresponded with the extent of missing data in the estimation sample. Sampling uncertainty was handled by bootstrapping (n = 2500 replications). ¹⁴ The 95% credible intervals reported the upper and lower 95% of bootstrapped values. The 95% credible intervals encompassed the upper and lower 95% of bootstrapped values. The joint distribution of incremental costs and incremental health outcomes was presented on a cost-effectiveness plane (Figure 1). ¹⁵ The probability that the Spotlight-AQ intervention is cost-effective was illustrated using a cost-effectiveness acceptability curve (Figure 2). ¹⁶

Figure 1.

Cost-effectiveness plane. Illustrates the joint distribution of the incremental costs and incremental QALYs from 2500 bootstrap replications. The diagonal line through the origin represents the NICE cost-effectiveness threshold of £30 000 per QALY gained. The distribution illustrates a high likelihood that the Spotlight intervention is cost-effective (proportion of points below the diagonal line).

Figure 2.

Cost-effectiveness acceptability curve. The estimated incremental cost-effectiveness ratio (£3132 per QALY gained) was below the threshold used by NICE to determine cost-effectiveness (£30 000 per QALY gained), therefore the intervention was cost-effective compared with current practice.

Results

Main Trial

A total of 98 adults with diabetes participated (n = 49 intervention). Three people were excluded from the analysis because they had no randomization code or missing allocation variable, leaving 95 participants for these analyses (Table 2).

Table 2.

Descriptive Characteristics.

Variable	Full sample (n = 95)	Intervention arm (n = 46)	Control arm (n = 49)
Age, mean (SE)	52.49 (1.70)	51.61 (2.31)	53.32 (2.53)
Male, % (n)	49% (46)	60% (28)	39% (19)
Type 1, % (n)	65% (62)	74% (34)	56% (28)
Baseline HbA1c, mean (SE) mmol/mol (%)	65 (8.1) (2.1)(0.2)	64 (8.0) (2.4)(0.2)	66 (8.2) (3.2)(0.3)

HCP Burnout

A total of 18 HCPs completed the Maslach Burnout Inventory questionnaire at baseline and study end. Burnout among HCPs was significantly reduced across all subscales after using Spotlight-AQ (P ≤ .001), with doctors experiencing greater improvements than nurses (Table 3).

Table 3.

HCP Burnout.

	Emotional exhaustion			Depersonalization			Sense of personal achievement
	B/line	F/U	Diff	B/line	F/U	Diff	B/line	F/U	Diff
Whole Group (n = 18)	11.29 (2-32)	8.88 (2-29	−2.41 (0 to −18)	6.41 (1-23)	4.8 (0-16)	−1.61 (0 to −12)	33.82 (13-46)	38.4 (21-49)	4.61 (−4 to 13)
Nurses (n = 8)	8.44 (2-21)	6.77 (3-14)	−1.67 (1 to −8)	4.0 (1-11)	3.33 (0-9)	−0.67 (1 to −3)	34.1 (13-46)	39.21 (21-44)	5.11 (0-13)
Doctors (n = 10)	14.5 (4-32)	11.26 (3-29)	−3.24 (5 to −18)	9.13 (1 to 23)	6.25 (1-16)	−2.88 (1 to −12)	33.50 (21-44)	38.13 (32-49)	4.63 (−4 to 13)

Health care professionals also reported improved communication and greater focus in consultations. Free text responses showed that prior to use of the Spotlight-AQ intervention, HCPs felt frustration at not being able to deliver the high-quality care they are capable of due to high absence rates, having to cover other colleagues, constant pressure of failure to meet targets, high DNA rates by patients and perceptions that patients simply do not listen to them or take their advice. Following use of the intervention, HCPs reported higher levels of satisfaction, particularly around the greater focus on solution-based care pathways, and improved understanding of the challenges participants face beyond glycemic levels.

Cost-Effectiveness

The primary outcome of the study was consultation duration. Although these data have been previously reported, in summary, participants in the intervention arm showed a 20% reduction in consultation duration (mean 6.1 minutes, range −0.5 to −15 minutes, P ≤ .001) compared with the control group, which showed no change or an increase in duration (−0.9 to +10 minutes). The secondary outcome of HbA_1c improved in the intervention group by a mean of 6 mmol/mol (range 0 to −30; 0.55%) versus the control group mean of 3 mmol/mol (range 0-8; 0.25%; P ≤ .001).

The Spotlight-AQ intervention increased costs by £13 (US$16) per patient; however, the shorter visit lengths offered a cost saving. Cost savings per visit were variable across health care providers, that is, primary care practitioners saw a mean cost saving of £5.88 (US$7.50) (range £4.62-£8.82 (US$5.89-11.25)), diabetes educator mean cost saving was £16.66 (US$21.26) (range £13.09 to £24.99 (US$16.70-31.88)) and endocrinologist-led visit mean cost savings were £25.76 (US$32.87) (range £20.24-£38.64 (US$25.82-40.56)).

The estimated incremental cost-effectiveness ratio (£3132 (US$3995) per QALY gained) was below the threshold used by NICE to determine cost-effectiveness (£30 000 (US$38 294) per QALY gained). Therefore, the Spotlight-AQ intervention was cost-effective compared with current practice. Figure 1 illustrates the joint distribution of the incremental costs and incremental QALYs from 2500 bootstrap replications. The diagonal line through the origin represents the NICE cost-effectiveness threshold of £30 000 (US$38 294) per QALY gained. The distribution illustrates a high likelihood that the Spotlight-AQ intervention is cost-effective (proportion of points below the diagonal line). There is considerable uncertainty in the direction of the QALYs gained and cost improvement over six months (proportion of points on either side of the Y-axis and X-axis, respectively).

Figure 2 is a cost-effectiveness acceptability curve to illustrate the probability of cost-effectiveness over a range of threshold values. At a threshold value of £30 000 (US$38 294) per QALY gained, there is a 68% probability that the Spotlight-AQ intervention is cost-effective.

Discussion

This study demonstrates that the Spotlight-AQ intervention can improve HbA_1c levels in people with diabetes, despite shorter consultations and reduced HCP burnout, including emotional exhaustion, depersonalization, and enhanced sense of personal achievement, all while being cost-effective. Furthermore, shorter visit lengths provide additional cost savings across all health care provider groups. The Spotlight-AQ intervention reduced unscheduled visits to the emergency departments. Such visits have a median cost ranging from $972 to $1499, contributing about 2% of monthly diabetes-related costs during follow-up.

Health care professional burnout is at an all-time high,^7,8 with recent warnings about the associated harms. ² Such burnout is also associated with lack of lack mental or physical energy, resulting in the completion of high priority tasks but neglect of so-called low priority tasks such as reassuring patients. ¹⁷

An anonymous survey of members of the American College of Surgeons ¹⁸ found that 8.9% of 700 surgeons reported concern they had made a major medical error in the last three months. Over 70% of those attributed the error to individual factors, with a large statistically significant inverse relationship with mental quality of life, including all three domains of burnout (emotional exhaustion, depersonalization, and personal accomplishment). Each 1 point increase in depersonalization was associated with an 11% increase in the likelihood of reporting an error and each 1 point increase in emotional exhaustion was associated with a 5% increase. The authors concluded that major medical errors reported by surgeons are strongly related to a surgeon’s degree of burnout. It costs between US$300 000 and US$500 000 to replace an HCP depending on experience. Burnout is detrimental not only to HCP well-being but also to patient care and health care costs. ¹⁹ Burnout is associated with medical errors, poor quality of care and low patient satisfaction. ¹⁹

Furthermore, Sui et al ¹⁹ report physicians who experience a high degree of burnout took four times the amount of sick leave as those with non-high burnout (P = .127). Similarly, Soler et al ²⁰ report that physicians experiencing high burnout had poorer job satisfaction, greater intention to change job, greater sick leave utilization as well as greater (ab)use of alcohol, tobacco, and psychotropic medication.

High-quality and patient-centered health care is dependent upon the well-being and safety of HCPs. Overrunning clinics take a toll on HCPs and deny them time to refocus between patient visits. The time-saving nature of the Spotlight-AQ intervention not only enables clinics to run on time, but also offers HCPs time to pause, refocus, and commit their full attention to their next patient with active listening and empathy.

The traditional medical model of health care is inadequate for supporting and treating long-term conditions, like diabetes. Implementing a biopsychosocial model within existing health care structures has been a significant challenge. While there is willingness to adopt a more holistic approach, skepticism remains about its real-world feasibility. This study assessing cost-effectiveness and HCP burnout, along with our previous research including underpinning the theoretical model of care ⁴ , pilot study, ²¹ and feasibility study, ²² demonstrates that it is indeed possible to routinely deliver biopsychosocial care for all people with diabetes. The reduction in burnout was seen despite Spotlight-AQ being used in only part of an HCPs working life. As such, the effects of the intervention on burnout could have been diluted by other aspects of work and the potential of the intervention may therefore be even greater.

The main strength of the study is that it is the first to demonstrate the cost-effectiveness of implementing a biopsychosocial model within routine care. Additionally, this approach leads to shorter, more focused consultations and reduces burnout among HCPs. One limitation of the study is that external variables may have contributed to changes in participants’ HbA_1c levels. The purpose of Spotlight-AQ is to clarify each individual’s priority concerns across biopsychosocial domains, enabling routine visits to address unmet needs that are usually overlooked. As such, this broader focus facilitates discussion beyond HbA_1c and glycemic measures and enables wider person-centered support to be offered. This aligns with recommendations from the World Health Organization (WHO), NICE, and the American Diabetes Association (ADA). The burnout data are limited by the relatively small number of participating HCPs. Another limitation was the early closure of recruitment due to COVID restrictions and staff demands at the participating NHS sites. A causal relationship between the intervention and the outcomes was not established. Future research is planned to adapt the existing Spotlight-AQ intervention for use with those with low literacy and poor access to digital technologies.

Conclusions

Spotlight-AQ has demonstrated cost-effectiveness in routine care, reducing consultation times and unscheduled visits while improving HbA1c levels and addressing previously unmet biopsychosocial needs. It was associated with reduced HCP burnout, including emotional exhaustion and depersonalization, and enhanced their sense of personal achievement, with doctors experiencing a greater reduction in burnout than nurses. Spotlight-AQ meets the NICE threshold for cost-effectiveness, benefiting both people with diabetes and HCPs.