[ Scientific Activity - Actividad Científica ] [ Brief Communications - Temas Libres ]
Effect of the application of the new diagnostic criteria of diabetes mellitus in the prevalence estimates of the disease and diagnostic level in the general population in Galicia (North West of Spain).
Muñiz Javier; Cordido Fernando; López Rodríguez Isidro; Castro Beiras Alfonso
Instituto Universitario de Ciencias da Saude
Hospital Juan Canalejo
La Coruña, Spain
Materials and Methods
Objective: To study the effect of the application of the new diagnostic criteria of diabetes mellitus on the prevalence estimates of the disease by sex and age group in the population aged 40-69 years old in Galicia (north-west of Spain).
Methods: A reanalysis is performed of a cross-sectional study in the general population of Galicia aged 40-69 years old that included 1275 participants. "Old criteria" are considered diabetes under treatment at the time of survey and/or fasting glycemia above 140 mg/dl; "new criteria" are considered diabetes under treatment at the time of survey and/or fasting glycemia above 125 mg/dl. Point prevalence and 95% confidence intervals (95% C.I.) with both sets of criteria are presented.
Results: Prevalence of diabetes rose from 7.5% (95% C.I.= 6.1-9.1) to 10.4% (8.8-12.2) with the new criteria. The proportion of non diagnosed diabetes mellitus rose from 22.2% with the old criteria to 37.8% with the new ones. This proportion is very stable in the different age and sex groups studied.
Conclusions: The new diagnostic criteria of diabetes mellitus make it necessary to increase previous prevalence estimates by 30-35% in the middle age (40-69 years). The high proportion of non diagnosed diabetic patients in the population and the prognosis implications of this underdiagnosis suggests the need to increase efforts to detect these patients in the population.
Introduction: Recently, a report was published by the Expert Committee on the Diagnosis and Classification of Diabetes Mellitus which established new criteria for the diagnosis of diabetes based on fasting glycemia levels1. The reevaluation of the criteria were done to 1) avoid discrepancies between cut off points based on fasting glycemia and those based on glycemia two hours after glucose load and 2) to encourage the use of a testing procedure that is easier and yet equally precise in the diagnosis of diabetes.
Cross-sectional studies on a disease that is common and of high resource consumption, directly and indirectly, as is diabetes mellitus, allows for a better administration of resources for preventive and therapeutic activities. To better reach this aim, the use of the new criteria will force us to review the prevalence estimates available to date.
The objective of this work is to evaluate what effect this modification will have on prevalence estimates of diabetes mellitus using the new criteria, as compared with the old, in different age (decades) and sex groups in the general population in Galicia (north-west of Spain), ages between 40-69, as well as the potential effect on the diagnostic level in this group of interest.Top
Materials and Methods: To achieve the objective described above, a new analysis will be done of the only study available performed on the general population in Galicia. The field work of the mentioned study was done between 1988 and 1989 and the methods and results have been previously published in detail2.
Briefly, two stages sampling was used. In the first step, nine counties were selected, with probability of selection proportional to the population of the counties. In the second step, stratified random sampling by age (decades) and sex was done. The electoral census (where every person older than 18 years old must appear) was used to select the participants. Of the 1,939 people invited to participate, ages between 40 and 69, 299 were census errors (due to death, change of address and emigration); of the 1,640 contacted, 1,275 participated in the study, making the total participation rate 78%. Subjects were weighed and measured, a short questionnaire was administered and the fasting glycemia in capillary whole blood was determined using a ReflotronÒ system that has a mechanism that separates plasma and does the determination in this fluid.
In this work, "old criteria" are defined as those used in the study previously published: diagnosis of diabetes upon examination and/or fasting glycemia equal or above 140 mg/dl (7.84 mM/l) prior to treatment. Two hour after glucose load values were not available from this study and, therefore, were not included in the estimates of disease prevalence. "New criteria", compared to the previous ones, are those suggested by the Expert Committee1: diagnosis of diabetes upon examination and/or fasting glycemia =126 mg/dl (7 mM/l) prior to treatment.
Prevalence estimates and 95% confidence intervals are presented with both criteria and are expressed as percentages. The absolute increase in the estimate (expressed as %) is the result of subtracting the old estimate from the new. This result is divided by the old estimate and multiplied by 100 to obtain the relative increase in the estimate.Top
prevalence estimates with both criteria and the absolute and relative increases using the
new criteria are shown in Table 1.
Table 2 shows the proportion of non diagnosed diabetics in the population by sex and age group with both criteria.
Discussion: As could not be otherwise, the application of the new criteria based on fasting glycemia produces an increase in the prevalence estimates of diabetes mellitus (table 1). The increase in absolute terms is around 3%, but with wide oscillations among age and sex groups (from 0.8 to 5.3%). The relative increase is large and fairly constant in the different sex and age groups: close to 30-35%, suggesting a normal distribution that is similar in the different groups. In women of advanced age, the relative increase is slightly greater, close to 50%. The significance of this particularity, if any, cannot be extracted directly from the data.
The percentage of non diagnosed diabetic patients with the new criteria has also increased, the estimate going from 22.2% to 37.8%. This increase has brought about a great uniformity in the proportion of non diagnosed diabetic patients in the population, independent of age and sex group: around 35-40%. Nevertheless, one must be cautious with these estimates since, as observed in table 2, they are based on very few subjects. On this matter, we are still better off than the U.S. where using the old criteria, it is estimated that approximately half of diabetics are not diagnosed3, but puts us farther from reaching our objective of achieving full diagnosis and it obligates us to make greater efforts in the detection.
This diversion from an ideal diagnosis situation and the importance of an early diagnosis suggests that there is a need for intensifying our efforts for early detection (and management) of the disease. It is important to let physicians know about the new criteria and even transmit this information to the population through opportunistic detection strategies used for other common metabolic disorders in the population4.
In conclusion, the results of this study suggest that previous prevalence estimates of diabetes mellitus have to be increased by 30-35% in adults 40-70 years old in our area. The elevated proportion of non diagnosed patients and the implications of this underdiagnosis compels us to reevaluate the strategies used in detecting patients. Application of these figures by the physician to his/her own practice may help to estimate at the individual level the work overload that the new criteria will pose on him/her to reach previous levels of quality in the management of these patients.
1. Report of the Expert Committee on the Diagnosis and Classification of Diabetes
Mellitus. Diabetes Care 1997; 20:1183-97.
2. Muñiz J, Hervada J, Juane R, López-Rodríguez I, Castro-Beiras A. Prevalence of diabetes mellitus in the population aged 40-69 years in Galicia, Northwest Spain. Diabetes Res Clin Pract 1995; 30;137-42.
3. Roman SH, Harris MI. Management of diabetes mellitus from a public health perspective. Endoc Metab Clin N Am 1997; 26:443-474.
4.- Ministerio de Sanidad y Consumo. Consenso para el control de la colesterolemia en España. Ministerio de Sanidad y Consumo, Madrid, 1991.
This work has been possible thanks to a research grant from the Dirección Xeral de Atención Especializada de la Consellería de Sanidade of the Xunta de Galicia.
Questions, contributions and commentaries to the Authors: send an e-mail message (up to 15 lines, without attachments) to firstname.lastname@example.org , written either in English, Spanish, or Portuguese.
This company contributed to the Congress: