Medical brain notes 35

  Obtaining and Preparing Samples for Analysis W hen   we   first   use   an   analytical   method   to   solve   a   problem,   it   is not   unusual   to   find   that   our   results   are   of   questionable   accuracy   or   so imprecise   as   to   be   meaningless.   Looking   back   we   may   find   that nothing   in   the   method   seems   amiss.   In   designing   the   method   we considered   sources   of   determinate   and   indeterminate   error   and   took appropriate   steps,   such   as   including   a   reagent   blank   and   calibrating our   instruments,   to   minimize   their   effect.   Why,   then,   might   a   carefully designed   method   give   such   poor   results?   One   explanation   is   that   we may   not   have   accounted   for   errors   associated   with   the   sample.   When we   collect   the   wrong   sample   or   lose   analyte   while   preparing   the   sample for   analysis,   we   introduce   a   determinate   source   of   error.

  The Importance of Sampling When   a   manufacturer   produces   a   chemical   they   wish   to   list   as   ACS   Reagent   Grade, they   must   demonstrate   that   it   conforms   to   specifications   established   by   the   Ameri- can Chemical Society (ACS). For example, ACS specifications for NaHCO 3   require that   the   concentration   of   iron   be   less   than   or   equal   to   0.001%   w/w.   To   verify   that   a production   lot   meets   this   standard,   the   manufacturer   performs   a   quantitative   analy- sis,   reporting   the   result   on   the   product’s   label.   Because   it   is   impractical   to   analyze the   entire   production   lot,   its   properties   are   estimated   from   a   limited   sampling.   Sev-   eral samples are collected and analyzed, and the resulting mean,  X – ,  and standard de-   viation,   s,   are   used   to   establish   a   confidence   interval   for   the   production   lot’s   true mean,   μ where   n   is   t

  Designing A Sampling Plan A   sampling   plan   must   support   the   goals   of   an   analysis.   In   characterization   studies   a sample’s   purity   is   often   the   most   important   parameter.   For   example,   a   material   sci- entist   interested   in   the   surface   chemistry   of   a   metal   is   more   likely   to   select   a   freshly exposed   surface,   created   by   fracturing   the   sample   under   vacuum,   than   a   surface   that has   been   exposed   to   the   atmosphere   for   an   extended   time.   In   a   qualitative   analysis the   sample’s   composition   does   not   need   to   be   identical   to   that   of   the   substance being   analyzed,   provided   that   enough   sample   is   taken   to   ensure   that   all   components can   be   detected.   In   fact,   when   the   goal   of   an   analysis   is   to   identify   components present   at   trace   levels,   it   may   be   desirable   to   discriminate   against   majo

  Where to Sample the Target Population Sampling   errors   occur   when   a   sample’s   composition   is   not   identical   to   that   of   the population   from   which   it   is   drawn.   When   the   material   being   sampled   is   homoge- neous,   individual   samples   can   be   taken   without   regard   to   possible   sampling   errors. Unfortunately,   in   most   situations   the   target   population   is   heterogeneous   in   either time   or   space.   As   a   result   of   settling,   for   example,   medications   available   as   oral   sus- pensions   may   have   a   higher   concentration   of   their   active   ingredients   at   the   bottom of   the   container.   Before   removing   a   dose   (sample),   the   suspension   is   shaken   to   min- imize   the   effect   of   this   spatial   heterogeneity.   Clinical   samples,   such   as   blood   or urine,   frequently   show   a   temporal   heterogeneity.   A   patient’s   blood   glucose   level,