This is an edited extract from a published paper on the screening process done by UWI researchers: Yvonne Ann Batson, Professor Surujpal Teelucksingh, Dr Rohan Maharaj, Dr Brian Cockburn, Dr Virendra Singh and Sasha Balkaran.
To read more of the paper, click here.
Diabetes mellitus is a group of metabolic diseases characterised by hyperglycaemia which result from defects in insulin secretion, insulin action or both. Currently there are more than 150 million people with diabetes worldwide and it is estimated that this number will double by 2025. More than 90% of diabetics have type 2 diabetes mellitus (T2DM). Most of this increasing burden of diabetes and its complications will be borne by developing countries. In Trinidad and Tobago 60 years ago, acute infectious disease was the main cause of morbidity and mortality. By the 1960s, the island was already experiencing one of the highest prevalence rates of T2DM in the Western Hemisphere. The prevalence now is estimated to be approximately 15-20%, which means that about 160,000 of the population are afflicted with diabetes. A study in 1968 by Poon-King et al showed that T2DM was rare in those aged<20 years but within the last decade a striking increase in the number of children with obesity and T2DM has been observed, reflecting a trend seen in North America, Europe and Asia. Pilot data fromthe Paediatric EndocrineClinic, Mount Hope, Trinidad support our clinical impression that there are many cases of undiagnosed T2DM in the young.
Given its prevalence within the snowballing obesity epidemic, earlier diagnosis and appropriate management is likely to be more cost-effective than treating complications that accrue from undiagnosed or under-treated disease, potentially adding many years of improved quality of life.
The objectives of this study spearheaded by Professor Surujpaul Teelucksingh were to assess the prevalence of diabetes mellitus in schoolchildren in Trinidad using mass urinary screening for persistent glycosuria and the cost-effectiveness of this exercise.Children from the island of Tobago were not screened in this study. As such it is an eye opener not just for Trinidad but also for the Caribbean region where obesity is getting to be epidemic, highlights the ‘diabesity’ problem encountered in adult life and fosters the metabolic syndrome.
A cross-section of primary and secondary schoolchildren at 415 schools in the seven school districts in Trinidad was surveyed during January to June 2009. Children were instructed to collect an early morning sample of urine at home (taken before breakfast) and bring it to school for glucose testing using the UriScanHurine strip.
This convenient strategy was chosen for the following reasons: to ensure that students collected urine in a safe and comfortable environment (their homes) and to delay the time lag between collection and testing—a fresh sample collected at home and presented soon after arrival at school was a practicable solution.
Urine was tested at schools by trained personnel according to the manufacturer’s instructions.Children with a positive result for the first sample were asked to provide a second sample for testing by different trained personnel. Those with consecutively positive samples were invited to the hospital for an OGTT. The OGTT was performed in the morning after an 8-hour fast.
This cross-sectional survey of schoolchildren in Trinidad found the prevalence of T2DM to be 10.4/100,000, of pre-diabetes to be 7.5/100,000 and of T1DM to be 1.5/100,000. This exceeds our worst expectations. We anticipated finding one positive glucose test for every 10,000 children screened but obtained an overall prevalence of 3.2/10,000. In Japan, where a similar screening method has been employed, the annual incidence of T2DM was only2.63/100,000 in 6-15 year-olds for the period 1974-2002.
The increase in T2DM in children is causally influenced by the alarming increase in childhood obesity which is now classified as an epidemic. This increase in obesity is also apparent in Trinidad and Tobago where there has been a significant increase in obesity over the last decade. A cross-sectional study by Batson et alin 2001 demonstrated a prevalence of 5% obesity in schoolchildren attending secondary schools in Trinidad and Tobago. By comparison, in an ongoing study one decade later, the observed prevalence of obesity was 15% (unpublished observations).
The rising prevalence of obesity, the increasing frequency of diagnosis of diabetes in children and the prospect of a more meaningful impact from early intervention make schoolchildren a prime target for screening. Recognition and awareness of the problem can halt adult diabetes and demote Trinidad’s ranking as the fifth ‘fattest’ nation globally.
Screening for diabetes in schoolchildren in Trinidad, West Indies
Yvonne Ann Batson, Surujpal Teelucksingh, Rohan Maharaj,Brian Cockburn, Virendra Singh, Sasha Balkaran
Departments of Clinical Medical Sciences and Paraclinical Sciences, The University of the West Indies, MountHope, and Department of Science and Agriculture, The University of the West Indies, St Augustine Campus,Trinidad and Tobago.
Background:
Although the epidemic of diabetes in adults is well established and documented, informationon the epidemiology of type 2 diabetes mellitus (T2DM) in children and adolescents in Trinidad and Tobagois limited. Anecdotal reports suggest an increase in children and especially adolescents with T2DMamongst clinic attendees.
Objective:
To assess the prevalence of diabetes mellitus in school children screened by mass urinarytesting in Trinidad and to determine the effectiveness and cost-effectiveness of this screening method.
Design and Methods:
During January to June 2009, a cross-sectional survey was undertaken in schoolchildren aged 5–17 years, in the island of Trinidad only. Children were instructed to collect an early morning, pre-breakfast urine sample at home and to bring it to school for testing for glycosuria. Those withan initially positive result underwent a second urine test. Those with two consecutively positive urine testswere invited to undertake an oral glucose tolerance test.
Results:
67,000 (53.6%) children from a school-age population of 125,000 were screened. Twenty three werefound to have persistent glycosuria and 21 agreed to undergo an oral glucose tolerance test. Eight fulfilled theAmerican Diabetes Association’s criteria for diabetes and five for pre-diabetes. Of the eight with confirmed diabetes (aged 7–18 years), one (male, age 12 years) was slim and ketotic and required insulin for control. Of theother seven, five were overweight (three females, two males) with a BMI .85th per centile and two females wereobese (BMI.95th per centile). Five children (four females, one male) aged 12–14 years were diagnosed as prediabetic.
There is a prevalence of 10.4/100,000 schoolchildrenwith T2DM, and 7.5/100,000 with impaired glucoseintolerance among school children in Trinidad. Urine screening had a positive predictive value (PPV) of 65% fordetecting T2DM in schoolchildren. The cost of screening 67,000 children was US$55,080, a per capita cost ofUS$1. The cost of finding one case was US$4286. Economic analysis revealed that investing to find one casetoday yields a net present value of .US$7000, representing a 63% saving.
Conclusion:
Despite the low PPV of urine glucose testing, our data support the view that mass screening ofschoolchildren in Trinidad for T2DM is both feasible and cost-effective.
Keywords:
Type 2 diabetes mellitus, Childhood, Mass screening, Urine, Health care cost
Introduction
Diabetes mellitus is a group of metabolic diseases characterised by hyperglycaemia which result from defects in insulin secretion, insulin action or both. Currently there are more than 150 million people with diabetes worldwide and it is estimated that this number will double by 2025. More than 90% of diabetics have type 2 diabetes mellitus (T2DM). Most of this increasing burden of diabetes and its complications will be borne by developing countries. In Trinidad and Tobago 60 years ago, acute infectious disease was the main cause of morbidity and mortality. By the 1960s, the island was already experiencing one of the highest prevalence rates of T2DM in the Western Hemisphere. The prevalence now is estimated to be approximately 15–20%, which means that about 160,000 of the population are afflicted with diabetes. A study in 1968 by Poon King et al. showed that T2DM was rare in those aged, 20 years but within the last decade a striking increase in the number of children with obesity and T2DM has been observed, reflecting a trend seen in North America, Europe and Asia. Pilot data fromthe Paediatric Endocrine
Clinic, Mount Hope, Trinidad support our clinical impression that there are many cases of undiagnosed T2DM in the young.
Several studies have implied that there is an accelerated risk of retinopathy and nephropathy in youths with T2DMcompared with those with type 1diabetes mellitus (T1DM). One longitudinal studydemonstrated the earlyprogression of complications15 years after T2DMhad been diagnosed—as many as 15% had developed serious complications or died.
Earlier recognition and aggressive treatment of diabetesis likely to yield better outcomes than silent progression and ultimate presentation with complications.
Given its prevalence within the snowballing obesity epidemic, earlier diagnosis and appropriate managementis likely to be more cost-effective than treatingcomplications that accrue from undiagnosed orunder-treated disease, potentially adding many yearsof improved quality of life. Various methods areavailable for population screening. Glucose metabolismcan be evaluated reliably by the oral glucosetolerance test (OGTT) which also identifies earlier but equally important forms of glucose intolerance,vizimpaired fasting glycaemia (IFG) and impairedglucose tolerance (IGT), both of which are prediabeticconditions with a high risk of progression todiabetes. These forms of pre-diabetes are potentiallyamenable to intervention with lifestyle changes and/or medication.
Findings from several lifestyle intervention programmes, including the Diabetes PreventionProgram, the Finnish Diabetes Study and The DaQuing IGT and Diabetes Study, show that it is possibleto prevent progression from pre-diabetes to diabetes byencouraging modest weight reduction. These dataprompted us to undertake mass screening of schoolchildrenusing simple, painless, sensitive urine tests whichif persistently positive could indicate diabetes.
Formore than three decades, Japan has maintained aprogramme of screening schoolchildren for glycosuria.
Between 1974 and 2004, over 9.2 million were screened:236 were diagnosed with diabetes, 84% of whom were obese, and the overall incidence was 2.55/100,000. Theincidence in ‘juniors’ (aged 13-15 years) was higher than in primary schoolchildren (aged 6-12 years), 6.27compared with 0.75/100,000, respectively.
The objectives of this study were to assess theprevalence of diabetes mellitus in schoolchildren inTrinidad using mass urinary screening for persistentglycosuria and the cost-effectiveness of this exercise.
Children from the island of Tobago were notscreened in this study.
Subjects and Methods
A cross-section of primary and secondary schoolchildrenat 415 schools in the seven school districts inTrinidad was surveyed during January to June 2009. Children were instructed to collect an earlymorning sample of urine at home (taken beforebreakfast) and bring it to school for glucose testingusing the UriScanHurine strip.
This convenient strategy was chosen for thefollowing reasons: to ensure that students collectedurine in a safe and comfortable environment (theirhomes) and to delay the time lag between collection and testing—a fresh sample collected at home andpresented soon after arrival at school was a practicablesolution.
Urine was tested at schools by trained personnelaccording to the manufacturer’s instructions.
Children with a positive result for the first samplewere asked to provide a second sample for testing bydifferent trained personnel. Those with consecutivelypositive samples were invited to the hospital for anOGTT. The OGTT was performed in the morningafter an 8-hour fast. Blood samples were drawn forfasting glucose and then 120minutes following theoral load, as described by the World HealthOrganization.
The follow-up report on the Diagnosis of DiabetesMellitus was used to define diabetes mellitus (DM) and pre-diabetes respectively, i.e. fasting bloodglucose >7.0 mmol/L or 2-hour post-glucose load>11.1 mmol/L and IFG 5.6–6.9 mmol/L or IGT 2-hour post-glucose load .7.8–11.0 mmol/L.
Permission was obtained from the Ministries ofEducation and Health, and from the EthicsCommittee of The University of the West Indies, StAugustine Campus. Parental permission for screeningwas a pre-requisite for entry into the study.
Results
125,000 children from a school population of 250,000were invited to participate in the study; 67,000 wereeventually screened. The overall response rate was53.6%: 74% in primary schools and 39% in secondaryschools. Twenty-three children with persistent glycosuriawere identified, an overall incidence of 34.3/100,000. Twenty-one of these 23 consented to glucosetolerance tests and eight (38%) of them wereconfirmed as diabetic: one type 1(male, aged 12)and seven type 2 (five females aged 6-15 years andtwo males aged 16 and 17). Five (24%) of the 21 (fourfemales, one male) were diagnosed as pre-diabetic(24%). Eight children exhibited normal glucosetolerance. The remaining two children underwentOGTTs elsewhere and were advised that their resultswere negative for DM. These latter ten children areclassified as having persistent glycosuria.
The data demonstrate a prevalence of 10.4/100,000for T2DM, 1.5/100,000 for T1DM, and 7.5/100,000 for IGT in schoolchildren in Trinidad and Tobago.
Urine screening had a PPV of 65% for detectingT2DM in schoolchildren. The cost of screening 67,000 children was US$55,080, a per capita cost ofUS$1. The cost of finding one case was US$4286.For any rational investment, expenditure has to beweighed against potential gains. The objective of ourcost analysis was to provide aconservative estimateof expected future expenditure compared with theopportunity cost of screening now. Limb amputationis the least expensive and least likely recognised complication and therefore provides the mostconservativeestimate of future costs.
The estimated cost of treating a diabetic septic foot in the future is conservatively estimated at US$16,000which, if discounted at 4.25% (the Trinidad and Tobago/US$ market rate), gives a current cost of US$11,710,which is far greater than the opportunity cost ofidentifying one diabetic child today. Therefore investingto find one case todayyields a net present value (presentvalue of treating the least costly complication minus thecost of finding one case) of more than US$7000,representing a 63% saving (Table 1).
Discussion
This cross-sectional survey of schoolchildren inTrinidad found the prevalence of T2DM to be 10.4/100,000, of pre-diabetes to be 7.5/100,000 and ofT1DM to be 1.5/100,000. This exceeds our worstexpectations. We anticipated finding one positiveglucose test for every 10,000 children screened butobtained an overall prevalence of 3.2/10,000. InJapan, where a similar screening method has beenemployed, the annual incidence of T2DM was only2.63/100,000 in 6-15 year-olds for the period 1974-2002.
This is the first attempt to screen children inTrinidad for diabetes mellitus (DM). The project hasalready positivelyaffected the lives of several childrenand their families through necessary medical intervention.
The mother of one of the children diagnosedwith DM also had an OGTT done and was alsodiagnosed with T2DM. The grandmother of a 7-year-oldchild with a normal OGTT was also newlydiagnosed with T2DM. There werechildren withnormal glucose tolerance tests who were persistentlyglycosuric but whose OGTTs were negative usingcurrent criteria. This may be a special group whorequire close scrutiny because many of them havestrong family histories of T2DM. In this regard,recent evidence suggests that, in children, currentlyacceptable ormoglycaemic levels between 4.7 and5.4 mmol/L might indicate pre-diabetes or predictdiabetes in later life.
The increase in T2DM in children is causallyinfluenced by the alarming increase in childhoodobesity which is now classified as an epidemic. Thisincrease in obesity is also apparent in Trinidad andTobago where there has been a significant increase inobesity over the last decade. A cross-sectional studyby Batson et al. in 2001demonstrated a prevalenceof 5% obesity in schoolchildren attending secondaryschools in Trinidad and Tobago. By comparison,in an ongoing study one decade later, the observedprevalence of obesity was 15% (unpublishedobservations).
An unexpected but fortuitous discovery was thatmany of these children’s parents also had diabetes butwere unaware of it until their children were studied. Achange in lifestyle to include physical activity andhealthy dietary habits is essential to halt thisemerging trend of obesity and T2DM in our youngpeople. These data support the need for formalscreening protocols to detect undiagnosed T2DM inthe young. The relatively high prevalence ofundiagnosedchildhood T2DM in Trinidad coupled with acost-effective screening method justifies mass screeninghere and in similar environments. While primaryprevention of diabetes in childhood is likely to be the most cost effective measure, secondary prevention through screening, early identification and treatmentof those alreadyaffected adds value by allaying costly complications that can accrue if left undiagnosed.
The sensitivity and specificity of urine screeningvaries between studies. A sensitivity of 35.3% and aspecificity of 99.7% were reported by West andKalbfleisch.26 In a prospective diagnostic studycollecting urine from fasted subjects, Davidsonet al.27 reported a sensitivity of 74% and a specificityof 86% whereas Friderichsen and Maunsbach reported a sensitivity of 20.8%, specificity of 99.14%and a PPV of 46.88%. We are aware that urinalysis isnot sufficiently sensitive to detect a high proportion ofpersons with diabetes and has the potential to findmorefalse negatives than blood screening, but, nevertheless,these mass screening data suggest that, in a developingcountry such as Trinidad and Tobago, childhooddiabetes of all types can be identified by glycosuria.
This method is more acceptable as it is easy to perform,painless, inexpensive, safe and specific for detectingasymptomatic children. The overall response rate was53.6%. The response rate in primary schools was 74%compared with 39% in secondary schools. It isunfortunate that we obtained a lower response rate insecondary schools which could have biased ourfindings. This is the age group with the higherprevalence of the disease and hence our data couldhave under-estimated the scale of the problem. Wesubsequently learned that the relatively low responserate in secondary schoolchildren was related to fearsthat it might have been a surreptitious exercise to screenfor illicit drug use or pregnancy. Such misapprehensionsshould be allayed by informing and educating subjectsbefore screening is undertaken.
The rising prevalence of obesity, the increasingfrequency of diagnosis of diabetes in children and theprospect of a more meaningful impact from earlyintervention make schoolchildren a prime target forscreening.
The impact of this project will continue to be feltfor years as the benefits to children (affected andunaffected) become ever more widely recognised inthe community.
In the last two decades, T2DM, a disease usuallydiagnosed in adults, has been shown to be anemerging problem in children and adolescents inboth developing and developed countries, and it isbecoming a serious public health concern. This studyhas demonstrated that urinalysis is a cost-effectivescreening tool, despite its poor PPV and sensitivity.
The relatively high prevalence of undiagnosed childhoodT2DM uncovered by the survey in Trinidadand Tobago, coupled with this cost-effective methodof screening, justifies this type of mass screening here and in similar environments.
Acknowledgment
The work was supported by the Helen BhagwansinghDiabetes Education, Research and PreventionInstitute (DERPI) and The University of the WestIndies, St Augustine, Trinidad, West Indies. |
