Original Research | Published: 17 June 2024

Real-world clinical validation of the Qatar pre-diabetes risk score: a cross-sectional study


Request reuse permissionopen-url
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) licenseopen-url


Introduction Pre-diabetes stands as a prominent, independent risk factor for the onset of type 2 diabetes (T2D), with 5%–10% of individuals with pre-diabetes progressing to T2D annually. The effectiveness of rigorous lifestyle interventions in averting the transition from pre-diabetes to T2D has been substantiated across multiple investigations and populations. Consequently, the clinical imperative of early pre-diabetes detection becomes unequivocal. This study assessed the validity of the recently developed pre-diabetes risk score in Qatar (PRISQ) in a real-world clinical setting.

Research design and methods We recruited 1021 walk-in participants from 3 different health centres of Qatar’s Primary Health Care Corporation. Only adult people without known pre-diabetes or diabetes were included in the study. Along with blood collected for the haemoglobin A1c (HbA1c) test to confirm pre-diabetes, we recorded the age, gender, weight, waist circumference, systolic and diastolic blood pressure, nationality, smoking state and family history of diabetes. Negative predictive value, positive predictive value, sensitivity and specificity of PRISQ were computed.

Results Of the 1021 participants, 797 agreed to provide blood. HbA1c test revealed that 21.9% of the 797 subjects had pre-diabetes (HbA1c between 5.7% and 6.5%) while 3.3% had undiagnosed diabetes (HbA1c≥ 6.5%). Using a PRISQ cut-off of 16, PRISQ sensitivity exceeded 90% in all subgroups of individuals aged 40 years and above, regardless of ethnicity. We did not see any significant improvement in PRISQ sensitivity when we considered the family history of diabetes.

Conclusions We confirmed a good PRISQ diagnostic rate for pre-diabetes from a representative sample of the Qatar population recruited in a real-world clinical setting. PRISQ can potentially play a significant role in curbing the T2D epidemic sweeping Qatar and beyond.

What is already known on this topic

  • The progression from pre-diabetes to type 2 diabetes (T2D) can be prevented through lifestyle intervention in a significant proportion of individuals. Timely diagnosis and the subsequent appropriate treatment are both needed to reduce the burden of T2D. Current pre-diabetes diagnosis, that is, fasting blood glucose, haemoglobin A1c or oral glucose tolerance test, involves an invasive and potentially expensive blood test, unsuitable for populations screening. Hence, urgent development of non-invasive tools for large-scale screening, especially in high pre-diabetes prevalence countries, is needed.

What this study adds

  • The current study provides evidence that a pre-diabetes risk score in Qatar (PRISQ) we developed recently using clinical data from Qatar biobank, and which uses only non-invasive parameters, can correctly diagnose pre-diabetes in more than 90% of ≥40 years individuals in a real clinical setting.

How this study might affect research, practice or policy

  • PRISQ stands as a powerful solution, providing a rapid and cost-effective pre-diabetes diagnosis through exclusively non-invasive parameters. Tailored for population screening in primary healthcare centres, its applicability transcends Qatar, holding the promise to globally address the T2D crisis. Furthermore, PRISQ could benefit populations living in regions where invasive and potentially expensive blood tests may not be easily accessible. The implementation of PRISQ in primary healthcare centres is not just beneficial; it is imperative for proactive healthcare, offering relief to healthcare systems grappling with the diabetes epidemic.


Pre-diabetes is a metabolic condition characterised by elevated blood glucose concentrations that surpass normal levels but do not meet the definitive diagnostic criteria for diabetes mellitus. Current estimates indicate that approximately 7.5% of the global population is affected by pre-diabetes.1 Importantly, pre-diabetes confers a substantial increase in the risk of progression to type 2 diabetes (T2D). Longitudinal studies have consistently demonstrated an annual transition rate of 5%–10% among individuals with pre-diabetes progressing to manifest T2D.2 Projections forecast a rise in the global pre-diabetes prevalence in the next 20 years,1 exacerbating the anticipated future incidence of T2D unless effective preventive measures are promptly implemented.

A fundamental quandary encountered when tackling pre-diabetes revolves around its inherent potential for prolonged asymptomatic manifestation, a trait it shares with T2D. Consequently, it can surreptitiously and progressively compromise vital organs, thereby establishing the groundwork for the onset of diabetes-associated microvascular and macrovascular complications.3

Epidemiological studies across diverse Middle Eastern nations, particularly within the Gulf Cooperation Council (GCC) countries, have documented a notable prevalence of pre-diabetes akin to T2D.4 5 Considering the previously mentioned annual transition ratios, these findings evoke heightened apprehension in the region. Encouragingly, a multitude of scientific findings have established that the progression from pre-diabetes to T2D can be averted, or at minimum delayed, in a substantial proportion of pre-diabetes individuals through interventions such as intensive lifestyle modification or pharmacological approaches like metformin.6–11 This observation highlights the paramount significance of promptly identifying individuals with pre-diabetes. This proactive measure holds substantial promise as a cost-effective intervention in mitigating the prevailing T2D epidemic sweeping the Middle East region.

According to a a secondary data analysis of a cross-sectional survey that included 2497 Qatari nationals aged 18–64 years, the prevalence of pre-diabetes was 11.9%.12 However, a pilot screening programme (n=3515) involving individuals at risk for T2D (eg, obese, hypertensive, family history of diabetes (FHD)) revealed that about 21% of the screened individuals had pre-diabetes.13 Other studies from the different nations of the GCC have reported pre-diabetes rates from 19.4% to 40 in Kuwait,5 14 25.5% in a cohort of Saudi adult males4 and 17.02% in adult Emirates.15 These rates are among the highest in the world.1 Given the high prevalence of pre-diabetes in Qatar, we have recently developed a pre-diabetes risk score (PRISQ) based on data collected from 6000 Qatari individuals sourced from Qatar Biobank.16 PRISQ uses only non-invasive parameters, including age, gender, body mass index (BMI), waist circumference (WC) and blood pressure, to generate a score indicative of the risk level of having pre-diabetes. PRISQ has a performance of 80% and a sensitivity of 86%.16 In pursuit of clinical validation of PRISQ, with the prospect of deploying this tool in primary healthcare clinics, we embarked on a study to ensure the tool’s robustness, reliability and accuracy when applied in real-world medical scenarios.

Research methods

Study setting

The study was conducted in Primary Health Care Corporation (PHCC), Qatar’s largest primary care provider. Approximately 70% of Qatar’s population is registered with PHCC across 31 health centres located in the country’s three geographical regions (Central, Western and Northern).

Study population and design

A meticulously designed cross-sectional study was conducted to validate the efficacy of PRISQ within a real-world clinical setting. Individuals aged 18 or above with no previous history of diabetes mellitus and who could communicate in English or Arabic were eligible for inclusion. Exclusion criteria included pregnancy, not being able to consent, having a BMI<18.5 or >60 kg/m2, and having an existing diagnosis of either T1D or T2D. The recruitment took place from December 2022 to March 2023.

For the recruitment of participants, we used a probability sampling method. One health centre was randomly selected from each geographic region. Every fifth individual was approached to participate in the study at each health centre.

The sample size required for the study was calculated as 176 subjects with pre-diabetes and 528 without pre-diabetes. According to the American Diabetes guidelines,17 pre-diabetes subjects were defined as those individuals with Haemoglobin A1c (HbA1c) between 39 mmol/mol (5.7%) and 47 mmol/mol (6.4%), whereas controls were those with HbA1c<39 mmol/mol (5.7%). An additional 20 participants in each group were included to factor withdrawals. Therefore, the overall sample required was estimated to be 744. The calculations were performed according to the equations in Flahault et al.18

Participant recruitment

Trained data collectors approached potential participants as they entered waiting areas of health centres. They established the eligibility of participants, following which they obtained written informed consent before proceeding to data collection.

Study locations and data collection

The study was conducted at one randomly selected PHCC health centre from each of the three geographical regions of Qatar as follows—Central Region (Al Thumama Health Center), Western Region (Muaither Health Center) and Northern Region (Al Gharrafaat Al Rayyan Health Center).

Potential participants were provided with information about the study, and if they agreed to participate, written informed consent was obtained. Data collectors recorded the following information from participants: age, gender, nationality, weight, height, WC, systolic and diastolic blood pressure readings, smoking status, and FHD. BMI (kg/m2) was calculated by dividing the weight in kilograms by the square of the height in metres. The WC measurement was performed using a flexible measuring tape wrapped around the waist at the midpoint between the iliac crest and subcostal margins. The tape was placed directly on the skin or a thin layer of clothing without any compression. Additionally, a certified phlebotomist meticulously drew blood to facilitate subsequent HbA1c analysis.

PRISQ model

PRISQ was initially developed in a randomly selected training set (1902 subjects with pre-diabetes and 3912 subjects without pre-diabetes(80% of the total sample)) and internally validated in a testing set (983 without pre-diabetes and 471 with pre-diabetes (20%). The clinical data used to develop PRISQ were obtained from the Qatar Biobank cohort, which enrols people from the general population.19 PRISQ includes five predictors: age, gender, BMI, WC, and systolic and diastolic blood pressure. The PRISQ score ranges from 0 to 45, and at a cut-off point of 16 yielded sensitivity and specificity of 86.2% (95% CI 82.7% to 89.2%) and 57.9% (95% CI 65.5% to 71.4%), respectively. The area under the curve (AUC) of the score was 80% (95% CI 78% to 83%).

Statistical analysis

All statistical analyses were performed using R V.4.0, RStudio V.1.2.5031 (R Core Team (2021)), and Stata V.16 (StataCorp). Participants were divided into subgroups according to their age in years (all, ≤40 or>40) and nationality’s geographical region. The analysis conducted in our study included two distinct risk levels to assess the likelihood of having pre-diabetes, as determined by the PRISQ score. We categorised the participants into two groups: those with a PRISQ score ≤16 are considered to have low risk level of having pre-diabetes, are thus free of pre-diabetes and coded 0 while those with a PRISQ score >16 are considered to have increased risk level of having pre-diabetes and coded 1.

The total score for each participant was calculated by summing the scores attributed to each variable category by multiplying the coefficient by 10 and rounding to the nearest integer following the logistic regression equation in our previous paper.16 Predicted values, obtained based on the PRISQ score of each participant and observed values, obtained based on the HbA1c value measured from the blood of each participant, were used to generate confusion matrices and calculate the true and false positives (FPs) and negatives, which are defined as fellow20 21:

  • True positives (TP) are the cases where PRISQ correctly predicted the positive class.

  • True negatives are the cases where PRISQ correctly predicted the negative class.

  • FPs are the cases where PRISQ incorrectly predicted the positive class when the actual class was negative.

  • False negatives are the cases where PRISQ incorrectly predicted the negative class when the actual class was positive.

The positive predictive value (PPV), which measures the probability that a positive test result is truly positive, and the negative predictive value (NPV), which measures the probability that a negative test result is truly negative, were calculated using the following formulas:

Display Formula

Display Formula

The sensitivity, which measures the proportion of actual positive cases that the test correctly identifies as positive, was calculated as follows20 21:

Display Formula

The specificity, which measures the proportion of actual negative cases that the test correctly identifies as negative, was calculated as follows:

Display Formula

We used the  Inline Formula  Test to compare the sensitivities of PRISQ in different subgroups. Statistical significance was considered at p<0.05.

To generalise PRISQ across various ethnicities within the Qatari population, we stratified the cohort into distinct subgroups based on participants’ citizenship, categorised following established global regional demarcations.22

Patient and public involvement

Volunteer participants were not involved in any stage of the research process.


Study population description

The chart of the flow of the study is shown in figure 1. The baseline characteristics of the study population are summarised in table 1. Among the 1021 individuals who initially consented to participate, only 816 agreed to donate blood for the HbA1c test. Following the exclusion of individuals with a BMI>60 and BMI<18.5 kg/m2, 797 participants (65.11% of whom were female and had an average age of 37±10 years) were kept in the final cohort. Consequently, PRISQ validation was evaluated within this subset.

Figure 1
Figure 1

Flow chart of the study design. BMI, body mass index; HbA1c; haemoglobin A1c; NPV, negative predictive value; PPV, positive predictive value; PRISQ, pre-diabetes risk score in Qatar.

Table 1
Baseline characteristics of the participants

Among the 797 individuals, 541 (67.87%) exhibited overweight or obesity, defined as BMI ≥25 kg/m². Within the male subset of 278 participants, 107 possessed waist circumferences (WC) below 94 cm, 93 had WC between 94 and 102 cm and 78 had WC beyond 102 cm. Among the 519 female participants, 134 had WC less than 80 cm, 122 ranged from 80 to 88 cm, and 263 exhibited WC greater than 88 cm. Hypertension, defined as systolic blood pressure (SBP) ≥130 mm Hg or diastolic blood pressure (DBP) ≥80 mm Hg, was observed in 51.7% of males and 37.3% of females. Current smokers represented only 8% of the sample; thus, we did not use this trait in further analysis.

Validation of the diagnostic accuracy of PRISQ

Of the 797 individuals, 596 (74.8%) were without pre-diabetes, 175 (21.9%) had pre-diabetes and 27 (3.3%) had undiagnosed diabetes (figure 2A). Pre-diabetes and diabetes were defined based on HbA1c values.

Figure 2
Figure 2

(A) Prevalence of pre-diabetes and diabetes in the whole cohort (n=797). (B) Levels of risk based on PRISQ score in the whole cohort. (C) Prevalence of pre-diabetes and diabetes in the group of individuals younger than 40. (D) Prevalence of diabetes and pre-diabetes and diabetes in individuals gaed 40 and above. Pre-diabetes and diabetes were defined based on HbA1c values according to the American Diabetes Association (ADA) criteria as described in methods. PRISQ, pre-diabetes risk score in Qatar.

In the subsequent analysis, we evaluated the discriminative efficacy of PRISQ by calculating standard diagnostic test metrics, specifically sensitivity, specificity, PPV and NPV. To ensure the comprehensive detection of individuals with hyperglycaemia, encompassing pre-diabetes and diabetes, we adopted a threshold criterion. Specifically, individuals were categorised according to their PRISQ scores. Those scoring a score of 16 or below were assigned to the low-risk category while scores ranging from 17 to 27 indicated a moderate risk of harbouring pre-diabetes or diabetes. Individuals with scores exceeding 27 were classified as high-risk individuals. Employing these delineations, we observed that 420 individuals, constituting 53% of the total cohort, possessed PRISQ scores surpassing 16, thereby implicating moderate to high-risk levels of pre-diabetes or diabetes, as depicted in figure 2B.

In the initial phase, we assessed the sensitivity of the PRISQ for the entire cohort, yielding a sensitivity of 70.3%. Subsequently, we conducted separate evaluations for the Qatari participants (sensitivity of 70.5%, n=297) and non-Qatari participants (sensitivity of 70.2%, n=500).

Given that age has a major contribution to the final score of PRISQ16 and considering that T2D typically manifests after the age of 40 years, we examined the prevalence of pre-diabetes in individuals younger than 40 years (40−) and those aged 40 years or older (40+) (figure 2C,D). Our analysis revealed that 35% of individuals in the 40+ cohort exhibit pre-diabetes, compared with only 20% in the 40− cohort.

The contingency tables pertaining to the entire sample population and various subpopulations are illustrated in online supplemental figures S1–S4.

Compared with PRISQ sensitivity in all participants and 40− groups (whole cohort (40−), Qataris (40−) and non-Qataris (40−), the PRISQ sensitivity in all the 40+ subgroups demonstrated a notable improvement, exceeding 90% (table 2). For instance, the PRISQ sensitivities in 40+ Qatari and non-Qatari subjects are, respectively, 93.9% and 96.3% compared with 42.9% and 54% in 40− Qataris and 40− non-Qataris, and 70.5% and 70.2% in all Qataris and all non-Qataris subjects (table 2). These differences in sensitivity were statistically significant (figure 3).

Table 2
PRISQ metrics in different subgroups
Figure 3
Figure 3

Comparison of PRISQ sensitivities between different subgroups. The values above the bars indicate sensitivity. ***p<0.001. PRISQ, pre-diabetes risk score in Qatar.

As the initial development of PRISQ was predicated on a cohort composed of 6000 Qatari Citizens, we assessed the generalisability of PRISQ across various ethnicities within the Qatari population. Prominent representation was observed in regions encompassing North Africa, southern Asia, southeastern Asia and Western Asia. The enrolment of individuals originating from other regions, such as Europe or North America, fell short of a substantive threshold required to form statistically significant subpopulations. As delineated in table 2 and figure 3, the discerned sensitivity of PRISQ among diverse ethnicities surpasses 90% within the 40+ age group, as compared with their 40− counterparts, where it attains a maximal value of 66.7%, explicitly observed within the North African subgroup. Figure 3 shows that the differences between 40+ and 40− sensitivities are statistically significant in all ethnicities.

The established body of literature affirms the pivotal role of an FHD as a prominent prognostic factor for the onset of diabetes. Consequently, we sought to examine whether the PRISQ sensitivity could be improved if we consider FHD. As delineated in table 2 and illustrated in figure 3, the discerned sensitivity of PRISQ in the 40+ Qatari individuals with an FHD showed a modest increment in comparison to when FHD was not considered (94.4% as opposed to 93.9%, p=0.67). In the 40+ non-Qatari with an FHD, PRISQ sensitivity reached 100%, demonstrating an improvement of nearly four percentage points compared with scenarios where FHD was not considered. However, this improvement is not significant (100% vs 96.3%, p=0.51)


As several studies have consistently shown, pre-diabetes can be reversed when met with targeted lifestyle interventions,6–11 underscoring the substantial potential for averting the progression to full-blown diabetes by timely identification and intervention in individuals with pre-diabetes. By capitalising on the reversibility of pre-diabetes and the cost-effectiveness of lifestyle interventions, a significant reduction in the burdens associated with T2D in terms of healthcare expenditures and immeasurable societal costs could be attained.

Recently, we developed a PRISQ using data from 6000 Qatari adults who participated in the Qatar Biobank cohort. PRISQ uses five non-invasive parameters (age, BMI, SBP, DBP and WC) and yields a score ranging from 0 to 45.16 A score above 16 indicates a moderate risk level of having pre-diabetes, while a score above 27 indicates a high-risk level.16 The sensitivity of PRISQ was 86%.16 In comparison to pre-diabetes risk scores formulated for adult populations in various international contexts, PRISQ exhibited superior discriminative capacity. As assessed by the AUC, PRISQ yielded an AUC of 80%, surpassing several analogous scores developed in disparate locales. Notably, these include risk scores from Indonesia (AUC: 62.3%),23 China (two distinct scores with AUCs of 74% and 70%)24 25 and the USA (AUC: 74%).26 Furthermore, our risk score demonstrated commendable sensitivity, quantified at 86.02%, with a designated cut-off value of 16.

In this study, we evaluated the performance of PRISQ in a real-world clinical setting by recruiting walk-in adult participants in three primary healthcare centres. Contrary to the original cohort, composed exclusively of Qatari citizens, in this investigation, we sought to ascertain the PRISQ performance across a diverse cohort, transcending ethnic boundaries.

Our results unequivocally affirm the outstanding performance of PRISQ, irrespective of participants’ ethnicity. PRISQ correctly identified more than 90% of the subjects with pre-diabetes among participants aged 40 years or above. This robust performance underscores the test’s reliability in capturing pre-diabetes in a clinical context, thus substantiating its potential as a pivotal tool in healthcare settings in Qatar to fight the T2D epidemic.

In the context of the relentless surge in the global incidence of pre-diabetes and T2D, particularly in the Middle East,27 28 a compelling imperative emerges to advance and enhance screening tools meticulously designed to enable comprehensive, population-wide screening. This need pertains notably to countries characterised by a heightened prevalence of these conditions, such as Qatar and other Gulf countries.28 29 In view of the annual conversion rate from pre-diabetes to T2D (5%–10%2), the critical significance of early diagnosis of pre-diabetes, and T2D for that matter, to formulate intervention strategies and the efficacious management of this widespread health challenge is undeniable. At a cut-off of 16, the sensitivity of PRISQ exceeded 90% in the 40+ subgroups regardless of ethnicity/race. This observation aligns with the fact that age significantly contributes to PRISQ score. Indeed, the age scores of 12 and 22 points for individuals between 36 and 55 years and those aged 55 or above, respectively. It is also noteworthy that T2D usually develops after the age of 40 years, and it is, therefore, not surprising that PRISQ performs better in the 40+ groups. We have observed a significant decline in the specificity of PRISQ, decreasing sharply from 66% in the under-40 group to 14% in the 40+ group. This steep decline is likely due to the higher prevalence of pre-diabetes in the 40+ cohort compared with the under-40 cohort (35% vs 20%). Additionally, the small sample size may contribute to this observed reduction in specificity. A wealth of evidence demonstrates the variation in diabetes prevalence among various racial and ethnic groups.30 31 Still, there is a lack of thorough knowledge of the underlying mechanics. The general view among the scientific and medical communities is that racial genetic variables have a negligible impact on the development of diabetes.32 Instead, a person’s susceptibility to pre-diabetes and T2D is more significantly influenced by non-genetic factors, which include physical characteristics, dietary habits, lifestyle decisions and more general social determinants of health like socioeconomic status, level of education, food insecurity and residential environment.33 Noteworthy is that all these non-genetic risk factors are amenable to cost-effective preventative interventions. Of particular interest is the observation that, aside from age and gender, the remaining predictors (BMI, WC, SBP and DBP) used in the computation of PRISQ could be modified, a characteristic that may elucidate the commendable performance of PRISQ across various ethnic and racial backgrounds.

The FHD exhibits a robust association with the disease’s incidence, yet the precise determinants underpinning this relationship remain incompletely elucidated.34 35 In the initial development of PRISQ, we consciously opted against incorporating FHD as a predictive variable, primarily due to the inherent challenges associated with this aspect. Many individuals may lack accurate knowledge of their family members’ medical histories,36 rendering familial history potentially unreliable.37 Moreover, situations may arise where individuals’ relatives have T2D or pre-diabetes but remain undiagnosed. Our current analysis has revealed that the differences between sensitivities of PRISQ within the 40+ age subgroups with and without FHD fail to attain statistical significance. This observation substantiates the rationale for our initial exclusion of FHD from the PRISQ model. The predictors used for PRISQ computation prove to be comprehensive and objective and can be readily ascertained by a healthcare professional in a primary healthcare setting, such as a nurse in a triage room.

Clinical importance of the study

With lifestyle interventions showing high efficiency in resolving pre-diabetes in many affected individuals, the significance of timely diagnosis of pre-diabetes cannot be overstated in the battle against the T2D epidemic sweeping the world in general and the Middle East region in particular. In this crucial endeavour, the PRISQ emerges as a beacon of hope. By relying solely on non-invasive parameters, the PRISQ tool offers a powerful means of diagnosing pre-diabetes quickly and cost-effectively. Importantly, contrary to blood tests, PRISQ is suitable for population screening in primary healthcare centres as the required parameters can all be quickly and non-invasively obtained by a nurse in the triage room before the encounter with the attending physician who is empowered to exercise discretion regarding the necessity of further recourse to confirmatory blood tests. This pragmatic approach necessitates the recognition that a nominal investment in preemptive measures to address potential FPs is financially judicious, given the far greater economic burden and clinical repercussions incurred by inadvertent oversight of TP cases. The potential of PRISQ impact extends far beyond Qatar’s borders, as it holds the potential to mitigate the worsening T2D crisis not only within the nation but also across the broader Middle East region, given the cultural, behavioural and ethnical similarities between the Middle Eastern populations, especially the nations of the GCC. The results we obtained with residents of Qatar originating from outside the Middle East countries could extend the use of PRISQ in other countries, especially from Southern, Southeastern and Western Asia. Embracing PRISQ in primary healthcare centres could be critical in preventive healthcare, promising to safeguard the health and well-being of countless individuals while relieving the strain on healthcare systems grappling with the diabetes epidemic.

The main limitation of this study was the relatively small number of participants and the lack of validation in populations outside Qatar. There is also the issue of low discordance between HbA1c and OGTT to diagnose pre-diabetes.38 39 It would have been ideal to use OGTT to confirm pre-diabetes. However, this approach would not be scientifically sound since PRISQ was developped based on HbA1c vlaues. Moreover, while we recognise that our diagnostic test may have low specificity, meaning it might produce some FPs, we are acutely aware of the potential consequences of overlooking TP results. FPs can lead to additional testing, which may incur limited additional costs, but missing TP cases can result in delayed intervention, disease progression and increased healthcare costs in the long term. Therefore, despite the limitations of our test, we prioritise sensitivity to ensure that genuine cases are not overlooked, even at the risk of some FPs. PRISQ can indeed be a starting point for further blood tests, especially OGTT, to confirm or reject the presence of pre-diabetes. We believe, however, that the strength of the study resides in the use of data from individuals from different ethnicities, indicating that PRISQ could potentially be useful in other populations.