SQL-从交叉表查询获取总计

Question

TRANSFORM Count(Research.Patient_ID) AS CountOfPatient_ID
SELECT DateDiff("yyyy",[DOB],Date()) AS [Age Group 22 - 24]
FROM (Research INNER JOIN Demographics ON Research.Patient_ID = Demographics.ID) INNER JOIN [Letter Status] ON Research.Patient_ID = [Letter Status].Patient_ID
WHERE (((DateDiff("yyyy",[DOB],Date())) Between 22 And 24))
GROUP BY DateDiff("yyyy",[DOB],Date())
PIVOT [Letter Status].Letter_Status;

This code lists individual ages, but I want to calculate the Total of all ages that fall within that range. So if there were two 22 year olds, and one 23 year old, the Total would be 3, as opposed to seeing all of their individual ages in the column...

Please help.

Answer 1

Please be sure you understand what you're getting when you use this method to determine a person's age.

DateDiff("yyyy",[DOB],Date())

If DOB = #1987-12-31#, on today's date (#2011-6-16#), DateDiff will give you this as the age:

? DateDiff("yyyy", #1987-12-31#, #2011-6-16#)
 24 

(And actually for any DOB during 1987, DateDiff with today's date would give you 24 as the "age". Furthermore, for any DOB during 1987, using that method on any day in 2011, you would get 24.)

OTOH, as age is commonly understood, one might say "Nonsense! His birthday is still more than 6 months away. He's only 23 today."

The reason for this discrepancy is that DateDiff("yyyy" only evaluates the year component of two dates. Consider these two dates which are one day apart:

? DateDiff("yyyy",#2010-12-31#,#2011-1-1#)
 1

That same thing happens when you use that expression to calculate "age".

To return the age as commonly understood, you can use an expression like this:

? DateDiff("yyyy", #1987-12-31#, Date())+ _
            Int( Format(Date(), "mmdd") < Format( #1987-12-31#, "mmdd"))

I copied that expression from the Access Web: Calculate Age of a person . That page includes other approaches for determining age.

None of this matters if, apparently like the OP, you want all DOB in 1987 to be counted as age 24 today, or on any day in 2011.

Answer 2

You need to define the age ranges in a switch statement. To make it easier I've created a inline query to calcuate the age (I'm assuming its on the research table).

TRANSFORM Count(Research.Patient_ID) AS CountOfPatient_ID
SELECT  SWITCH (Age <= 22, "Under 22",
        Age > 22 and AGE <= 24 , "Between 22 And 24",
                Age > 24 and AGE <= 26,  "Between 24 And 26",
                Age > 26 , "Over 26") as Age_Range
FROM 
   (Research
    INNER JOIN (SELECT ID,  DateDiff("yyyy",DOB,Date()) as AGE
        FROM Demographics) Demographics
    ON Research.Patient_ID = Demographics.ID) 
    INNER JOIN [Letter Status] 
    ON Research.Patient_ID = [Letter Status].Patient_ID

GROUP BY  AGE


PIVOT [Letter Status].Letter_Status;