To rule out or confirm presence of diabetes, following symptoms should suffice for a clinical investigation :-
i. When a person eats too much and too often/
ii. If he passes large quantity of urine at short intervals.
iii. If he drinks too much water, at short intervals
iv. If he experiences dryness of mouth, inspite of drinking large quantity of water.
v. If there is pruritus on and around genitals.
vi. Rough and dry skin.
If the said symptoms are prominent, it should be taken as an indication of probable presence of sugar. So, as a first step, go in for a routine urine examination and if there is even 0.5% sugar in urine, then, to make doubly sure, get the urine examined twice after pause of a day, followed by a third test on the third day. If all the clinical tests point out to presence of urine, then, have your blood tested for presence or otherwise of sugar (glucose) in blood. First blood sugar test should be on empty stomach, followed by another test after taking meals (that is, test should be carried after 1-2 hours of taking meals). Positive results will confirm diabetic status of the patient.

In certain cases, sugar level may rise only after taking meals but may be normal/near normal on fasting level. In such a case, after careful adherence to diet control, P.P.sugar may also normalize. In borderline cases, (in fasting or P.P.test results), one or two more tests may be had with marginal cut in sugar intake. Where, there is incidence of rapid rise or fall in blood sugar readings, G.T.T. test should be carried to determine as to after which time the sugar level rises/falls.

Confirmed diabetics should, at regular intervals (say at least once in a month) get their blood tested and adjust their diet, exercise and drug according to advice & guidance of their doctors.

If tests reveal presence of sugar in urine but absence thereof in the blood, then it should be viewed more seriously, as detailed earlier. We repeat that mere or occasional presence of urine sugar is not a case of ‘true’ or ‘confirmed’ diabetes, unless there is presence of sugar in blood and the latter one is a true sign of ‘true and confirmed’ Diabetes Mellitus.

As for self-tests, it can only be pointed out that such tests are neither precise nor are they reliable as they merely point out, if at all, to near about indications only and thus, are no substitutes for actual blood sugar tests, carried out by a reliable and dependable laboratory.

Finally, le the test be carried out at two /three different laboratories and result should be matched. If the results are in reasonable proximity to each other, there is no reason for authenticity of test results, but high difference in results, is bound to create confusion which should be dispelled. Tests carried out in big public hospitals are often found to be unreliable and inaccurate, for the simple reasons that, due to messing up of hundreds of samples (specimens) chances of wrong labeling, erratic tests and inaccuracy cannot be rule out. It is not that an accusing finger is raised towards public hospitals which have to carry out large number of tests, hence chances of ‘human failure’ or a ‘slip’ cannot be ruled out, due to rush of work.