# US 7,465,559 B2 · Method for calibrating antibiotic disk diffusion testing of microorganisms > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 7,465,559 B2 - Document ID: US7465559B2 - Status: Granted - Kind: B2 - Application number: 10/474,488 - Filed: 2002-04-08 - Published: 2004-09-09 - Granted: 2008-12-16 - Earliest claimed: 2001-04-10 - Assignees: Bioscand AB - Inventors: Goran Kronvall - Web page: https://patentgrep.com/patent/US7465559B2 ## CPC - C12Q 1/18: Testing for antimicrobial activity of a material - C12Q: MEASURING OR TESTING PROCESSES INVOLVING ENZYMES, NUCLEIC ACIDS OR MICROORGANISMS (immunoassay G01N33/53); COMPOSITIONS OR TEST PAPERS THEREFOR; PROCESSES OF PREPARING SUCH COMPOSITIONS; CONDITION-RESPONSIVE CONTROL IN MICROBIOLOGICAL OR ENZYMOLOGICAL PROCESSES ## Abstract A method for calibrating antimicrobic susceptibility testings of microorganisms comprises the steps of creating a histogram with a high response side and a low response side of isolates from a microbial species, which may contain resistant strains against an antimicrobial agent, calculating from the high response side of the histogram at least one statistical parameter, and defining a limit for susceptibility interpretation and comparative analysis of antimicrobic resistance, which is based on the statistical parameter. This limit separates susceptible strains from resistant strains against an antimicrobial agent. ## Claims ### Claim 1 A method for calibrating antimicrobic susceptibility testing data of microorganisms, wherein the method comprises the steps of creating a histogram based on in vitro susceptibility test data against an antimicrobial agent for isolates from a microbial species, which microbial species isolates may contain unknown resistant strains against said antimicrobial agent wherein the number of isolates of microbial species or percentage of the total number of isolates of microbial species is on a y-axis of said histogram and registered response values against said antimicrobial agent on a x-axis of said histogram; determining a position of a response peak of the histogram between a high response side and a low response side corresponding to isolates of microbial species susceptible to said antimicrobial agent; calculating from said high response side of said histogram at least one statistical parameter; and defining a limit for susceptibility interpretation and comparative analysis of antimicrobic resistance, which is based on said at least one statistical parameter; said limit separating susceptible strains from resistant strains against said antimicrobial agent. ### Claim 2 (depends on 1) Method as in claim 1, wherein said isolates arc clinical isolates. ### Claim 3 (depends on 1) Method as in claim 1, wherein said at least one statistical parameter is statistically obtained from a probability distribution calculated from said high response side of said histogram. ### Claim 4 (depends on 1) Method as in claim 1, wherein said microbial species is a bacterial species or a fungal species. ### Claim 5 (depends on 1) Method as in claim 1, wherein said antimicrobial agent is an antibiotic. ### Claim 6 (depends on 1) Method as in claim 1, wherein said antimicrobic susceptibility testing is a Minimal Inhibitory Concentration test, said registered response values being minimal inhibitory concentrations. ### Claim 7 (depends on 1) Method as in claim 1, wherein said antimicrobic susceptibility testing is a disk diffusion test, said registered response values being inhibition zone diameter values from paper disks impregnated with said antimicrobial agent. ### Claim 8 (depends on 3) Method as in claim 3, wherein said probability distribution is a Gaussian distribution. ### Claim 9 (depends on 8) Method as in claim 8, wherein said at least one statistical parameter is the mean, defining a position of the statistical response peak between a high statistical response distribution side and a low statistical response distribution side, and/or standard deviation. ### Claim 10 (depends on 9) Method as in claim 9. wherein said limit is defined as three times said standard deviation below said mean. ### Claim 11 (depends on 9) Method as in claim 9, wherein said Gaussian distribution is a standard Gaussian distribution. ### Claim 12 (depends on 9) Method as in claim 9, wherein said Gaussian distribution is a normalized Gaussian distribution of a theoretical normal population. ### Claim 13 (depends on 12) Method as in claim 12, wherein said normalized Gaussian distribution is construed from said position of said statistical response peak, said high statistical response distribution side, and its mirror image. ### Claim 14 (depends on 12) Method as in claim 12, wherein a total number for said theoretical normal population is construed by a procedure comprising the steps of calculating a slope for a line through said number of isolates of adjacent registered response values starting on said high response distribution side; detecting a shift in slope direction, which represents the position of said statistical response peak, thereby defining said mean; and calculating said total number as the doubled sum of half the number of isolates at said position plus the number of isolates having higher response values. ### Claim 15 (depends on 13) Method as in claim 13, wherein said normalized Gaussian distribution is construed graphically. ### Claim 16 (depends on 14) Method as in claim 14, wherein said normalized Gaussian distribution is determined by a procedure comprising the steps of calculating for each registered response value a percentage value as percent of said total number; calculating accumulated percentage values for each of said number of isolates or percent isolates on said statistical high response distribution side; performing a linearized transformation of said accumulated percentage values against said statistical response values; and calculating the equation constants for a linear relationship obtained from said transformation; thereby defining said normalized Gaussian distribution of said theoretical normal population. ### Claim 17 (depends on 15) Method as in claim 15, wherein said graphical construction is performed by means of an undirected graphical model or a directed graphical model. ### Claim 18 (depends on 16) Method as in claim 16, wherein said calculation of equation constants is performed by means of the least squares method. ### Claim 19 (depends on 16) Method as in claim 16, wherein said calculation of accumulated percentage values is performed in the direction 100% to 0%. ### Claim 20 (depends on 16) Method as in claim 16, wherein said transformation k a probit transformation. ### Claim 21 (depends on 20) Method as in claim 20, wherein said standard deviation is calculated from said probit transformation.