In the prac­tice of med­i­cine, get­ting it right is chal­leng­ing. Or per­haps bet­ter to say get­ting it right all the time is dif­fi­cult. In med­ical school, you are taught how to take a thor­ough his­to­ry from a pa­tient and then per­form an ex­am­i­na­tion and come up with some form of plau­si­ble di­ag­no­sis.

Un­for­tu­nate­ly, no­body strolls in­to the of­fice and ir­refutably states that they have stage four ovar­i­an can­cer.

As a doc­tor, you are meant to lis­ten care­ful­ly to your pa­tient, gath­er the im­por­tant de­tails, sift out the un­nec­es­sary in­for­ma­tion, and like a de­tec­tive, fo­cus on the clues and put the pieces of the puz­zle to­geth­er.

More of­ten than not, af­ter all of this, there may not be a def­i­nite an­swer, and then we must re­sort to med­ical tests or in­ves­ti­ga­tions to achieve an ac­cu­rate di­ag­no­sis.

Many of these tests are stan­dard and com­pul­so­ry. For in­stance, a pelvic ul­tra­sound scan is used to di­ag­nose uter­ine fi­broids or a blood test for anaemia.

How­ev­er, the dilem­ma in med­ical test­ing lies in when is it too lit­tle or in­deed too much.

Un­der­test­ing can lead to a de­lay in di­ag­no­sis or treat­ment, while in­ap­pro­pri­ate tests or over-test­ing wastes fi­nite health re­sources.

Nev­er­the­less, I am the first to ad­mit that I may “over-in­ves­ti­gate,” with the ob­vi­ous ex­cuse be­ing that work­ing in on­col­o­gy has made me fair­ly sus­pi­cious that every­one who walks through my door has can­cer un­til proven oth­er­wise.

Of course, I ex­ag­ger­ate. But take ovar­i­an can­cer as an ex­am­ple.

This gy­nae­co­log­i­cal can­cer is woe­ful­ly un­der-de­tect­ed, as with women wait­ing up to six to nine months to ar­rive at con­firmed di­ag­no­sis, the ma­jor­i­ty of them will, un­for­tu­nate­ly, have ad­vanced dis­ease by that time.

There are sev­er­al rea­sons for this. Ovar­i­an can­cer has very vague symp­toms, from tum­my bloat­ing to change in ap­petite or weight, feel­ing full quick­ly or some neb­u­lous pain.

These non-spe­cif­ic symp­toms can mean any­thing and of­ten oc­cur in be­nign con­di­tions.

Hence, it may take some time be­fore some­one con­sid­ers a gy­nae­co­log­i­cal ori­gin and then does the ap­pro­pri­ate in­ves­ti­ga­tions for this.

I have used ovar­i­an can­cer as a prime ex­am­ple of the pit­falls of un­der-test­ing and the im­por­tance of ap­pro­pri­ate­ly as­sess­ing every­one who presents with a con­cern.

How­ev­er, this is all for ar­gu­ment’s sake.

The re­al­i­ty in every­day med­ical prac­tice is that most women who do have gen­er­al com­plaints of pain or bloat­ing will NOT have ovar­i­an can­cer. In fact, they will not have can­cer at all.

Know­ing who to in­ves­ti­gate, how much to in­ves­ti­gate and when to stop in­ves­ti­gat­ing if you come up with noth­ing, is chal­leng­ing.

Un­sur­pris­ing­ly, most doc­tors ac­knowl­edge the cur­rent propen­si­ty for over-test­ing but freely ad­mit that it is now fu­elled by a cul­ture where we all like to feel re­as­sured, in ad­di­tion to the un­der­stand­able fear of mal­prac­tice claims if a prob­lem is missed.

There should be a judge­ment about ben­e­fit ver­sus risk. In sta­tis­ti­cal terms, over-test­ing does not in­crease sen­si­tiv­i­ty (the prob­a­bil­i­ty of de­tect­ing a dis­ease if it is present), but it does in­crease the num­ber of false pos­i­tives (ab­nor­mal test re­sults when there is, in fact, no dis­ease).

This may then lead to po­ten­tial­ly risky in­ter­ven­tion which is en­tire­ly un­nec­es­sary. For in­stance, if every­one with a headache had an MRI scan, most would not have a brain tu­mour, but the scan may re­veal in­ci­den­tal dis­cov­er­ies which may them­selves prompt avoid­able or even haz­ardous surgery.

You can then ar­gue it’s best not to know of these, thus spar­ing doc­tors from the al­most ir­re­sistible urge to fix what isn’t bro­ken!

I have al­so found that pa­tients of­ten in­cor­rect­ly equate the or­der­ing of tests with qual­i­ty and car­ing.

A rel­a­tive of mine show­ered praise on her doc­tor who or­dered an un­nec­es­sary MRI (about $6,000), which was fol­low­ing an un­nec­es­sary CT ($5,000), fol­low­ing an un­nec­es­sary blood test ($300), when all she re­al­ly need­ed was an over-the-counter lax­a­tive ($20) and words of re­as­sur­ance ($0).

In the end, it is al­ways about who is go­ing to ben­e­fit from a test and be­ing able to jus­ti­fy it us­ing the cor­rect ap­proach with the nec­es­sary ex­pla­na­tions, risks in­volved etc.

Some of this I have found comes with ex­pe­ri­ence. Ul­ti­mate­ly, when it comes to in­ves­ti­ga­tions, com­bin­ing some de­gree of clin­i­cal sus­pi­cion with good old-fash­ioned com­mon sense should pre­vail.