I recently became reacquainted with colleague Madeline Ducate, PhD, who supports pharmaceutical houses with the laboratory needs for international clinical trials. Madeline is occasionally confronted with a lack of laboratory reference intervals in under-resourced countries, where cultural norms or the lack of funding limit specimen collection and processing from healthy subjects.
In vitro diagnostics companies may provide in-house reference intervals in their package inserts for kits or controls, but hematology instrument distributors seldom provide normals with their equipment. I suggested to Madeline that for routine CBCs and coagulation tests she could apply textbook normals unless dealing with specialized populations such as sickle cell, thalassemia, or leukemia, however the pharmaceutical distributors are reluctant to use generalized reference intervals even when conducting large-scale screens. How would our participants address this issue?
To elaborate, medical laboratories are required as a matter of good laboratory practice to generate their own reference intervals, but I wonder how well this demand supports clinical decision making? Many tertiary care facilities pay med students, residents, nursing students, and medical lab science students $5 a tube for normal range specimens, then separate, aliquot, and freeze the plasmas or sera. Do these aliquots adequately represent the institutions' patient population demographics? What about normals for pediatric or geriatric patients? Others purchase frozen, or in the case of CBC parameters, preserved liquid materials from international distributors. Institutions with multiple sites attempt to compute a general reference interval for all sites. How closely do these match local patient mix?
Further, the more experienced clinicians, while they may compare their patients' results with their lab's reference interval, broaden their decisions based on prior experience, patient's general physical condition, current circumstances, or their own favorite references. Physicians in under-resourced countries interpret with no reference intervals at all. So, when publishing reference intervals, how finely should we slice the pie? If our lower HGB limit for maies is 13.7 g/dL and the Rodak textbook says 13.5, does this make a difference to the hematologist? What about therapeuitc targets, is a change from an INR of 2.5 to 2.8 important? Please respond with your opinion here.