The Bariatric Surgery Assessment and Prioritisation Tool (BAPT) prioritised those likely to gain greater benefits from surgery such as those who had diabetes for a shorter period and younger patients. Notably, patients’ diabetes improvement increased with their BAPT scores as did their health-related quality of life (HR-QoL) 12 months post-surgery. The study authors from the Queensland Health, Brisbane, Queensland, Australia, said these initial findings indicate that the BAPT is functioning as designed as a patient prioritisation instrument for bariatric surgery.
The authors explained that the BAPT was designed as a patient prioritisation instrument to assess patients with excessive weight and type 2 diabetes suitable for bariatric surgery. For this study, they assessed the development and pilot application of the BAPT to prioritise 292 patients referred from September 2017 to August 2019, 212 of whom underwent surgery between December 2017 and August 2020. The examined whether the instrument successfully identified those who gained the greatest benefits including weight loss, diabetes remission, reduction in comorbidities and HR-QoL.
The researchers considered the clinical characteristics of patients across the range of BAPT scores to determine whether the BAPT was able to prioritise those who should theoretically benefit the most. They then examined the patients’ clinical outcomes, stratified by their BAPT scores and HR-QoL, to consider whether high-scoring patients achieved better outcomes.
BAPT
BAPT identified five attributes, each of which contained various ‘levels’: impact on potentially reversible conditions (9 levels), duration of diabetes (3 levels), age group (4 levels), BMI category (3 levels) and surgical risk (3 levels). Overall, there was a maximum of 9 × 3 × 4 × 3 × 3 = 972 possible health states for the five BAPT attributes. The 1000Minds software reduced the mean number of choice tasks to a more-manageable 24 choice decisions. From the respondents’ decisions, the software calculated weights for each level within a criterion as well as the overall criterion weight.
Scores were normalised across criteria to provide relative preference weights, which are then scaled so the sum of all criteria ranges from 0 (least preferred) to 100 (most preferred); the possible range for the BAPT was 2–98 with higher scores indicating those who are predicted to benefit the most.
Outcomes
At referral, the cohort was on average (mean ± SD) 52 ± 8.7 years old and 57.1% female, with a mean BMI of 46.1±7.0 kg/m2; 21.9% were Aboriginal or Torres Strait Islanders. The average HbA1c was 8.77±1.5%; 99.7% of patients required oral medications for diabetes. All patients reported comorbidities, the most common of which were hypertension (86.3%), dyslipidaemia (85.2%) and sleep apnoea (66.0%).
The BAPT scores of 292 patients referred to the BSI were normally distributed between 12 and 78 (mean ± SD: 43±14). The distribution of BAPT scores was similar across patients who had surgery (42.5±13.2), those who remained under review (47.1±13.5) and those excluded from the service (42.0±14.2).
Referred patients were 22–66 years old and younger patients had higher BAPT scores overall - the average age of patients scoring 20–29 was 55.4±6.2 years; that of patients scoring 60–69 was 45.2±10.7 years (p<0.001). Nevertheless, 91% of the cohort that progressed from referral to surgery was older than 40 years and the average age of patients having surgery was 52.2±8.4 years.
Patients with higher BMI had higher BAPT scores (p<0.001) - patients scoring 20–29 had an average BMI of 41.3±5.5 kg/m2, while those scoring 60–69 had an average BMI of 50.7±6.9 kg/m2. Further, a greater proportion of those with higher scores had a BMI ≥40kg/m2.
The average HbA1c across all BAPT scores was similar at referral, although the lowest average HbA1c was seen in low-scoring patients (BAPT 10–19; 8.2±0.6%) and the highest occurred in high-scoring patients (BAPT 70–79; 9.5±1.7%), there was no correlation between HbA1c and BAPT score (p=0.73). However, the researchers reported a significant association between BAPT score and diabetes medication use pre-surgery.
The proportion of patients in each BAPT group who required insulin decreased with increasing BAPT score (p<0.001) - 75.8% of patients scoring 20–29 required insulin, compared to 48.1% of those scoring 60–69. BAPT scores were also higher for those with a shorter duration of diabetes before referral - 82% of those scoring above 70 had diabetes for less than four years, while those with diabetes for longer than eight years populated the lower-scoring groups.
Patients are also scored on the presence of selected comorbidities, which varied widely between BAPT score ranges and while the proportion of patients reporting weight-related joint pain and sleep apnoea increased at higher BAPT scores (joint pain: 38.5% at BAPT 20–29 v. 70.4% at BAPT 60–69; sleep apnoea: 47.1% at BAPT 20–29 v. 80% at BAPT 60–69), there were no trends regarding other comorbidities.
Of the 212 patients that had surgery, 130 patients had reached a minimum of 12 months post-surgery before data collection for the study stopped. The compared data for the 130 patients at 12 months post-surgery to their pre-surgery values and then stratified by their BAPT scores to consider whether those with higher scores had better outcomes (Figure 1).
Patients with higher pre-surgery weight and BMI had higher BAPT scores however, there was no significant difference in the percentage of weight lost (i.e. % BMI decrease) by patients with different BAPT scores (p=0.73). Differences were noted between BAPT scores regarding diabetes-related outcomes. While there was no significant difference in HbA1c values pre-surgery (p=0.36), higher-scoring patients had lower HbA1c at 12 months (p<0.01) and had a significantly greater decrease in HbA1c (p<0.01). For example, those with BAPT scores of 20–29 had an average pre-surgery HbA1c of 8.2±0.7%, which decreased to 7.3±1.8% post-surgery - an 11.8% improvement. In contrast, the pre-surgery average HbA1c of patients with BAPT scores of 70–79 was 9.5±1.8%, decreasing to 5.3 ± 0.4% - a 46.4% improvement, approximately four times greater than the BAPT 20–29 group.
Differences were also noted in patients’ diabetes medications after surgery. All patients required oral medications, insulin or both at referral. Post-surgery, 60–80% of patients across all BAPT scores discontinued insulin treatment, with no trend regarding BAPT score. However, a greater proportion of higher-scoring patients discontinued oral medications after 12 months (p<0.001). Overall, using the definition of remission of diabetes of HbA1c below 6.5% (48 mmol/mol) in the absence of glucose lowering medications, 39.3% of the sample obtained diabetes remission; this increased with BAPT score (p=0.327) with those who scored 50 or more were substantially more likely to obtain remission (57% vs 31%).
A significant difference was noted in the 18–49-year-olds compared to those older than 50 years, where the younger group achieved a greater decrease in average HbA1c (27.3% v. 20.1%, p<0.01).
Similar analyses of the results with BMI cut-points of 40, 45 and 50kg/m2 found a significant difference in BMI decrease at 12 months in patients with a starting value above 45 kg/m2. Their average BMI decreased by 25.6 ± 10.3% - a 3.3% greater decrease than the 22.3 ± 7.9% achieved by patients with BMI below 45 kg/m2. A lower BMI cut-point of 40kg/m2 found differences in the proportion of patients able to discontinue medications for diabetes by 12 months post-surgery.
In all three analyses, a greater proportion of the higher-BMI group discontinued all diabetes-related medications, though the difference was only statistically significant with a cut-point of 40kg/m2 (59.5% v. 32.0%, p<0.01). No significant differences in BMI or diabetes care were observed with a cut-point of 50kg/m2.
Those with low BAPT scores tended to have better HR-QoL pre-surgery than those with high BAPT scores. The change in AQoL-4D scores at 12 months post-surgery tended to be greater for those with higher BAPT scores (e.g. mean change in AQoL-4D: BAPT < 40 = 0.058±0.277 vs BAPT 40 and above = 0.174±0.235; p=0.095); this is a 10% vs 33% improvement in HR-QoL. There was no statistically significant difference in AQoL-4D scores across BAPT categories (p=0.184); the BAPT category 50–59 was the only category with a significant improvement in AQoL-4D score (0.200±0.246; p=0.032).
“In this pilot study, the BAPT successfully prioritised those who had diabetes for a shorter period and younger patients, who realised greater improvements in their diabetes, and those with higher BMI,” the authors concluded. “Most notably, patients’ diabetes improvement increased with their BAPT scores, as did their health-related quality of life 12 months post-surgery, indicating that the instrument prioritised those likely to gain greater benefits. Overall, these initial findings indicate that the BAPT is functioning as designed as a patient prioritisation instrument for bariatric surgery.”
The findings were featured in the paper, ‘Prioritising patients for publicly funded bariatric surgery in Queensland, Australia’, published in the International Journal of Obesity. To access this paper, please click here
Comments