Dissolution Testing Data and Equivalence Analysis

• Desktop program for dissolution testing staff at a pharmaceutical CMC research center
• One place for the whole run: import instrument data, calculate, store, compare, export a report
• Test conditions registered per project and reapplied to every run
• Results kept on a shared server rather than individual PCs, so past tests stay available

• Instrument readings retyped into spreadsheets, with formulas built by hand for each calculation
• A different output layout per instrument vendor, each needing its own handling
• Medium replaced after every withdrawal, so the correction accumulates with each sampling point and is easy to get wrong by hand
• Calculation files scattered across individual PCs, making past results hard to find and compare
• Equivalence criteria differing by formulation type, with the reasoning written up by hand each time

• Instrument files read as they are, separating standards from samples and picking up sampling time and vessel number from sample names
• Three output layouts recognized: Waters, Agilent UPLC, Agilent HPLC
• Three ways of handling standards: single-concentration standard, serial dilution, manually entered concentrations
• Corrections accumulated by sampling order, with uncorrected and corrected rates shown together
• Test conditions stored alongside the result, so the conditions behind a number are visible from the result itself

• Conditions registered as medium volume, replacement volume, sampling volume, dilution factor and standard handling
• Registered conditions copied and adjusted, instead of re-entering a run that differs only by medium

Flow of work

• Instrument data file → dissolution calculation → shared server → graph comparison → spreadsheet report
• Project → test conditions → calculated results, with results reached by project and by month stored
• Server credentials held in encrypted form

• Instrument layout identified from the report title in the file's first cell, so no vendor has to be chosen by hand
• Older file formats converted on read, leaving the file from the instrument untouched
• Calibration built from standard weights and peak areas, with slope, intercept and coefficient of determination kept with the result
• Samples from different runs drawn on one graph, with mean, standard deviation, minimum and maximum
• Legend, Y-axis range and deviation markers toggled to get the view wanted

• Equivalence assessed under separate criteria for immediate-release and for enteric-coated or extended-release products
• Similarity factor reported together with each criterion met or missed, leaving the path to the verdict visible
• Report exported as a spreadsheet holding measurements, dissolution tables, curves, verdicts and the decision flow

• Calculation variables with units, dissolution formulas and the decision flow built into the program
• Grounds for a number checked on screen rather than looked up elsewhere

• Calculation, storage and reporting finished inside one program, starting from the instrument file
• Manual entry and hand-built formulas removed, so the same conditions give the same result
• Results gathered on the server, letting runs in different media or different samples be overlaid directly
• Equivalence reasoning stored with the result and carried straight into the report
• Displayed fields and calculation handling revised over several rounds as requests came in from use