JP2009176023A - Program for converting examination value of medical examination - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a program for easily converting an examination value of a medical examination into a numerical value comparable with an examination value measured on a different measurement condition. <P>SOLUTION: A distribution type of the examination value before the conversion is decided based on previously acquired attribute information of the examination value before the conversion. When the decided distribution type is a normal distribution type, the numerical value having the same deviation value as that of the examination value before the conversion is calculated as the examination value after the conversion in a distribution of a target examination value, based on: the examination value before the conversion; a reference range of the examination value before the conversion; and a reference range of the target examination value. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a test value conversion program for centrally managing health check test values measured under different measurement conditions.

  As a result of the health check, the test value and the reference range (normal range) are usually written together for each test item. This reference range is a range that includes 95% of test values excluding the upper 2.5% and the lower 2.5% in the distribution of test values of healthy subjects. NCCLS (US National Committee for Clinical Laboratory Standards) ) Is set for each laboratory according to the calculation method defined by (see Non-Patent Document 1).

  The reference range of the inspection values varies considerably from inspection facility to inspection facility. For example, γ-GTP, which is one of the inspection items for biochemical inspection, has a reference range of 10 to 47 (IU / l) for one inspection facility, while a reference range of another inspection facility is 16 to 73. (IU / l). The reason for the variation in the reference range is that many measurement methods and analyzers exist for the inspection of the same inspection item, and the inspection value varies depending on the measurement conditions.

  It is said that it is preferable to judge the deterioration or improvement of the results of health examinations compared with the past values, but this variation in the examination values of the health examinations allows the subject to observe the changes in the examination values over time It is a hindrance. That is, when a subject receives a health check at a different inspection facility from the previous time, the subject cannot directly compare the obtained test value with the previous test value. In addition, even if a subject undergoes a medical examination continuously at the same inspection facility, if the inspection facility changes the measurement method or analyzer, it will not be possible to compare with past inspection results. .

Kunihide Ino, how to set the standard range, Gifu QA workshop report in 1995, p. 7-13

  In recent years, there has been a movement to standardize the measurement method of each inspection item and the specifications of the analysis apparatus in order to solve the above-described problems. However, clinical specimens are limited by the amount and location of collection, and sample components contain many mixed components, and it is necessary to measure a large number of samples in a short time from a single substance. There is a problem that it is difficult to unambiguously determine the optimal measurement method, and standardization has not progressed easily. In addition, even if standard measurement conditions are determined, in order to unify the measurement conditions at all inspection facilities, the financial burden on the inspection facilities accompanying the replacement of analyzers, etc. is large, and it has to be accumulated in the past. There is a problem that it becomes difficult to compare with the inspection value, which is not easy.

  On the other hand, as a method of converting measurement data measured under one measurement condition into a numerical value comparable to measurement data measured under a measurement condition different from the measurement condition, the correlation between the two measurement conditions using the same sample is obtained. A method of examining and converting measurement data by a conversion formula based on the correlation is generally performed (hereinafter referred to as a correlation method). However, when this correlation method is applied to the conversion of test values for medical examinations, it takes a lot of resources such as human resources, specimens, time, and reagents to determine the conversion formula. Therefore, it is not realistic to use the correlation method.

  The present invention has been made in view of the present situation, and an object of the present invention is to provide a program for easily converting a test value of a health check into a numerical value that can be compared with a test value measured under different measurement conditions.

  The present inventor has focused on the fact that the distribution type of the test value of the healthy person is the same even if the measurement conditions such as the measurement method and the analysis apparatus are different, after extensive studies for the purpose of solving the above problems, After trial and error, the present invention was completed.

  That is, the first invention is a process for converting the pre-conversion test value measured under one measurement condition into a post-conversion test value that can be compared with a target test value measured under a target measurement condition different from the measurement condition. A program to be executed by a computer, wherein the pre-conversion test value having a normal distribution type, the reference range of the pre-conversion test value, and the reference test value in the distribution of the target test value, the pre-conversion An inspection value conversion program for health check, which causes a computer to execute a normal distribution type conversion process for calculating a numerical value having the same deviation value as the deviation value of an inspection value as an inspection value after conversion.

  As described above, since the distribution type of test values does not depend on the measurement conditions such as the measurement method and the analyzer, if the distribution type of the test values before conversion is the normal distribution type, the distribution type of the target test values is also the normal distribution type. Become. In the normal distribution type processing in the present invention, as shown in FIG. 1, the deviation value of the inspection value before conversion in the normal distribution of the inspection value before conversion and the deviation value of the inspection value after conversion in the normal distribution of the target inspection value are the same. The post-conversion inspection value is calculated so that Thus, the post-conversion inspection value converted so as to have the same deviation value from the normal distribution to another normal distribution is a value comparable to the numerical value converted by the correlation method.

  Since the upper limit and lower limit of the reference range of the inspection value before conversion and the target inspection value are 2.5 percentile (lower limit) and 97.5 percentile (upper limit) in the distribution of each inspection value, the distribution type is normal. In the case of a distribution, it is easy to calculate the average value and standard deviation of each distribution from each reference range and calculate the deviation value of each inspection value. Therefore, according to the inspection value conversion program, as long as the reference range of the inspection value before conversion and the reference range of the target inspection value are given, the inspection value before conversion can be converted into the inspection value after conversion that can be compared with the target inspection value. The reference range of each test value is a value set by a predetermined method for each test facility as described above, and is general information that is also written in the test result of the health checkup. It can be easily obtained without. Therefore, according to the inspection value conversion program, the normal distribution type inspection value before conversion can be converted into the inspection value after conversion by a very simple method.

  Moreover, 2nd invention is the process which converts the test value before conversion measured on one measurement condition into the test value after conversion which can be compared with the target test value measured on target measurement conditions different from this measurement condition. A program to be executed by a computer, wherein a non-normalized distribution type inspection value, a pre-conversion inspection value reference range, and a target inspection value reference range are defined as a normal distribution type distribution. Each of the variables is converted so that the inspection value before conversion after the variable conversion, the reference range of the inspection value before conversion after the variable conversion, and the reference range after the variable conversion of the target inspection value are converted into the variables. In the converted distribution, a numerical value that is the same deviation value as the deviation value of the pre-conversion inspection value after variable conversion is calculated, and further, the numerical value is subjected to inverse conversion of the variable conversion and calculated as a post-conversion inspection value Distributed conversion processing computer Healthy diagnostic test value conversion program, characterized in that to execute.

  In the non-normal distribution type processing in the present invention, first, as shown in FIG. 2, each parameter is converted into a numerical value according to the normal distribution type by performing variable conversion of each parameter according to the distribution type of the inspection value before conversion. Thereafter, as shown in FIG. 3, as in the normal distribution type process, the deviation value of the pre-conversion test value in the normal distribution of the pre-conversion test value after variable conversion and the target test value after variable conversion The numerical value with which the deviation value of the inspection value after conversion in the normal distribution is the same is calculated, and the inspection value after conversion is calculated by further inversely converting the numerical value.

  Examples of distribution types that can be variable-converted into a form conforming to the normal distribution type include a log normal distribution type and a power distribution type. The lognormal distribution type can be converted into a normal distribution type by logarithmic conversion, and the power distribution type can be converted into a normal distribution type by power conversion. Since the lognormal distribution and the power distribution are significantly different from the normal distribution, conversion based on the deviation value cannot be appropriately performed as it is, but, as in the non-normal distribution conversion processing of the present invention, by variable conversion. If the distribution type is converted to a form that conforms to the normal distribution type and then converted based on the deviation value, even if the pre-conversion test value distribution type is not the normal distribution type, the conversion is equivalent to the value converted by the correlation method Post test values are obtained. Therefore, according to the inspection value conversion program, the pre-conversion inspection value that is a distribution type other than the normal distribution type can be simply and appropriately converted only by the reference range of the inspection value before conversion and the target inspection value.

  The third invention is a process for converting the pre-conversion test value measured under one measurement condition into a post-conversion test value that can be compared with a target test value measured under a target measurement condition different from the measurement condition. A program to be executed by a computer that calculates the estimated median value in the reference range of the target test value so that the ratio is the same as the ratio of the median value in the reference range of the test value before conversion. Is compared with the median of the test value before conversion, and if the test value before conversion is greater than the median value, the test value after conversion for the difference between the upper limit value of the reference range of the target test value and the estimated median value and the estimate Calculate the post-conversion test value so that the ratio of the median difference is the same as the ratio of the median difference between the upper limit of the reference range of the pre-conversion test value and the median difference. If the test value is smaller than the median, the target test value The ratio of the difference between the estimated median value and the converted inspection value to the difference between the estimated median value and the lower limit value of the reference range is the difference between the median value of the inspection value before conversion and the upper limit value of the reference range. A non-parametric conversion process for calculating a post-conversion test value that is the same as a value difference ratio is a health check test value conversion program.

  As shown in FIG. 4, the non-parametric conversion process according to the present invention performs conversion using the fact that the distribution of inspection values before conversion and the distribution of target inspection values are similar. This non-parametric conversion process performs conversion regardless of the type of distribution of the inspection value before conversion, so the distribution type of the inspection value before conversion is unknown or cannot be converted to a normal distribution. Even if there is, there is an advantage that the inspection value after conversion can be calculated. Moreover, a value close to a numerical value converted by the correlation method can be obtained by using the median value of the inspection value before conversion as a parameter instead of the distribution type. The median of the inspection value before conversion is a statistic obtained when the inspection facility sets the reference range, and is a numerical value that can be easily obtained.

  According to a fourth aspect of the present invention, a pre-conversion inspection value measured under one measurement condition is measured under a target measurement condition different from the measurement condition in a computer storing a distribution type list for the inspection value attribute. A program that executes a process of converting into a converted inspection value that can be compared with a target inspection value, from the attribute information of the pre-conversion inspection value acquired in advance, based on the distribution type list, When the distribution type determination process for determining the distribution type is executed, and when the distribution type determination process determines that the distribution type of the inspection value before conversion is a normal distribution type, the normal distribution type conversion process is executed by a computer, When it is determined in the distribution type determination process that the distribution type is a predetermined distribution type other than the normal distribution type, the non-normal distribution type conversion process is executed by a computer, and the distribution type determination process is performed. If the type is determined not in the predetermined distributed in a normal distribution type, a health diagnostic test value conversion program, characterized in that to execute the non-parametric conversion process on the computer.

  According to such an inspection value conversion program, it is possible to calculate the inspection value after conversion by an appropriate conversion method according to the distribution type of the inspection value before conversion. Since many of the distribution types for the inspection value attribute are described in the literature, the distribution type list can be created based on the information. Further, when there are a sufficient number of test values of healthy persons measured by the same measurement method, it is also possible to determine the distribution type of the test values from them and add it to the distribution type list.

  In the present invention, the inspection value attribute is attribute information for determining the distribution type of the inspection value, and the essential attribute information is the inspection item name. Some of the test items are known to have different test value distribution types depending on the gender of the subject. For these, the gender information of the subject is required as an attribute of the test value.

  As described above, according to the test value conversion program for health examination of the present invention, the post-conversion test value is calculated extremely easily compared with the correlation method by performing conversion based on the distribution type of the pre-conversion test value. be able to. Therefore, by using such a test value conversion program, the test results of medical examinees who have undergone health examinations at multiple test facilities can be converted into a comparable form, and the subject can grasp changes in test results over time. It becomes possible to do. In addition, by using such a test value conversion program, inspection facilities can convert past test result data into a form that can be compared with test values under different measurement conditions, so even if the measurement method or analyzer is changed, The facility can continue to centrally manage test results.

  Embodiments of the present invention are described according to the following examples.

  FIG. 5 shows an outline of the inspection value conversion system of this embodiment. The inspection value conversion system is configured by a computer in which an inspection value conversion program is installed. The computer captures inspection result data for each inspection facility, and converts each pre-conversion inspection value included in the inspection result data into a post-conversion inspection value that can be compared with a target inspection value measured according to a predetermined target measurement condition. Convert and output as conversion result data.

Table 1 is a distribution type list for the inspection value attribute stored in the storage device of the computer. The distribution type is uniquely determined by the inspection item, and the distribution type is determined by the gender of the subject in addition to the inspection item.

Table 2 is a list of reference ranges of target inspection values stored in the storage device of the computer. The reference range includes an inspection item set without distinguishing the subject, and an inspection item where the reference range is set for each gender of the subject.

  The inspection result data sent from the inspection facility includes inspection facility information, subject information, and inspection value information for each inspection item. The inspection facility information includes a facility ID, a reference range and a median value of each inspection item in the inspection facility. The subject information includes information such as the subject ID, the sex of the subject, and the age group. The inspection value information includes an inspection value before conversion for each inspection item associated with the subject information.

  The conversion result data output by the computer includes laboratory information, subject information, and conversion result information. The inspection facility information and the subject information are the same as the inspection result data. The conversion result information includes a pre-conversion inspection value and a post-conversion inspection value for each inspection item associated with the subject information. Furthermore, the conversion result information includes the reference processing range for the conversion processing method and the target measurement method for each post-conversion inspection value.

The inspection value conversion program of this embodiment causes a computer to execute the following processes.
・ Distribution type setting processing ・ Target standard range setting processing ・ Inspection result data import processing ・ Inspection value conversion processing ・ Conversion result data output processing

  The distribution type setting process is a process of acquiring and updating the latest distribution type list from the outside when a distribution type of inspection values for inspection items and sex is newly found by various investigations. The target reference range setting process is a process of acquiring and storing a list of reference ranges of the target inspection values from the outside. The inspection result data fetching process is a process for fetching the inspection result data sent from the testing facility and storing it in the storage device. Further, the inspection value conversion process is a process of converting the inspection value information included in the acquired inspection result data into a converted inspection value. The conversion result data output process is a process of outputting conversion result data including the post-conversion inspection value calculated in the inspection value conversion process.

  The control processing content of the inspection value conversion processing according to the main part of the present invention will be described with reference to FIG.

  In the test value conversion process, first, the distribution type of the pre-conversion test value is determined by comparing the inspection item name related to the pre-conversion test value to be converted and the gender of the subject with the list of distribution types (see Table 1). Determine (S1). Specifically, the program of the present embodiment corresponds to 12 types of distribution types such as a square distribution and a cubic root distribution as a normal distribution, a log normal distribution type, and a power distribution type. It is determined which of the distribution types corresponds to or does not correspond to any of them.

  After determining the distribution type of the inspection value before conversion, the process proceeds to step S2 to determine whether or not the determined distribution type is a normal distribution type. If the normal distribution type is determined, the normal distribution type conversion process (S4) is executed, the calculated inspection value after conversion is stored (S7), and the inspection value conversion process is terminated.

  On the other hand, if it is determined in step S2 that the distribution type is not normal, the process proceeds to step S3, and it is determined whether or not the determined distribution type is a predetermined distribution type. In the present embodiment, the non-normal distribution type and the ten power distribution types are set as the predetermined distribution types. If the determined distribution type or these correspond to these, the non-normal distribution type conversion process (S5) is performed. The calculated inspection value after conversion is stored (S7), and the inspection value conversion process is terminated. On the other hand, if it is determined that the determined distribution type is not the predetermined distribution type, the non-parametric conversion process (S6) is executed, the calculated converted inspection value is stored (S7), and the inspection value conversion process is terminated. To do.

  Next, the contents of the control process in the normal distribution type conversion process (FIG. 6, S4) will be described with reference to FIG.

In the normal distribution type conversion process, first, it is determined whether or not the following parameters exist (S11).
Xi: Inspection value before conversion
RL: Lower limit of the reference range before conversion (lower limit of the reference range of the inspection value before conversion)
RH: Reference range upper limit value before conversion (upper limit value of reference range for inspection value before conversion)
oRL: Lower limit value of target reference range (lower limit value of target test value reference range)
oRH: Target standard range upper limit value (Target test value standard range upper limit value)
Here, when it is determined that even one of the above parameters does not exist, the normal distribution type conversion process is terminated because conversion is impossible.

Next, the average value Xb and standard deviation SD of the inspection value Xi before conversion are calculated based on the reference range lower limit value RL and the upper limit value RH (S12, S13). Specifically, the average value XB and the standard deviation SD are obtained from the following mathematical formula.
XB = (RL + RH) / 2
SD = | RL−RH | / (2 × 1.96)

  Next, in the same manner as in steps S12 and S13, the average value oXB and standard deviation oSD of the target test values are calculated based on the target reference range lower limit value oRL and the upper limit value oRH (S14, S15).

Next, based on the average value XB of the inspection value Xi before conversion and the standard deviation SD, the standard deviation index ZXi of the inspection value Xi before conversion is calculated by the following formula (S16).
ZXi = (Xi−XB) / SD

Next, based on the standard deviation index ZXi, the average value oXB, and the standard deviation oSD, in the distribution of the target test value, the numerical value that becomes the same standard deviation index as the standard deviation index ZXi is calculated as the converted test value aXi, and the normal distribution The mold conversion process is terminated (S17). Specifically, the post-conversion inspection value aXi is obtained from the following mathematical formula.
aXi ＝ ZXi × oSD + oXB

  Next, the contents of the control processing in the non-normal distribution type conversion processing (FIG. 6, S5) will be described with reference to FIG.

In the non-normal distribution type conversion process, it is first determined whether or not the following parameters exist (S21).
Xi: Inspection value before conversion
RL: Lower limit of the reference range before conversion (lower limit of the reference range of the inspection value before conversion)
RH: Reference range upper limit value before conversion (upper limit value of reference range for inspection value before conversion)
oRL: Lower limit value of target reference range (lower limit value of target test value reference range)
oRH: Target standard range upper limit value (Target test value standard range upper limit value)
CD: Distribution type of inspection value before conversion When it is determined that there is no parameter, the non-normal distribution type conversion process is terminated because conversion is impossible.

Next, a conversion formula is selected based on the distribution type CD of the inspection value before conversion, and the above parameters Xi, RL, RH, oRL, oRH are converted to values cXi, cRL, cRH, coRL, coRH according to the normal distribution type. (S22). Specifically, as described below, logarithmic transformation is performed for the lognormal distribution type, and power transformation is performed for the power distribution type. In the following formula, p is an exponent in the power distribution type.
For lognormal distribution: Xi, RL, RH → cXi = ln (Xi), cRL = ln (RL), cRH = ln (RH)
Power distribution: Xi, RL, RH → cXi = Xi ^p , cRL = RL ^p , cRH = RH ^p

Next, the average value cXB and the standard deviation cSD of the inspection values before conversion after the variable conversion are calculated based on the variable-converted reference range lower limit value cRL and the upper limit value cRH (S23, S24). Specifically, the average value cXB and the standard deviation cSD are obtained from the following mathematical formula.
cXB = (cRL + cRH) / 2
cSD = | cRL−cRH | / (2 × 1.96)

  Next, in the same manner as in steps S23 and S24, based on the target reference range lower limit value coRL and upper limit value coRH after variable conversion, the average value coXB and standard deviation coSD of the target test value after variable conversion are calculated ( S25, S26).

Next, based on the average value cXB and the standard deviation cSD, the standard deviation index cZXi of the inspection value cXi before conversion after variable conversion is calculated by the following formula (S27).
cZXi = (cXi−cXB) / cSD

Next, based on the standard deviation index cZXi, the average value coXB, and the standard deviation coSD, in the distribution after the variable conversion of the target inspection value, the numerical value that becomes the same standard deviation index as the standard deviation index cZXi is the post-conversion inspection after the variable conversion The value caXi is calculated (S28). Specifically, the post-conversion inspection value caXi after the variable conversion is obtained from the following mathematical formula.
caXi = cZXi x coSD + coXB

  Finally, based on the distribution type CD of the pre-conversion inspection value, the post-conversion inspection value caXi after the variable conversion is inversely converted to calculate the post-conversion inspection value aXi (S29), and the non-normal distribution type conversion process is terminated. .

  Next, the contents of the control process in the nonparametric conversion process (FIG. 6, S6) will be described with reference to FIG.

In the non-parametric conversion process, first, it is determined whether or not the following parameters exist (S31).
Xi: Inspection value before conversion
RL: Lower limit of the reference range before conversion (lower limit of the reference range of the inspection value before conversion)
RH: Reference range upper limit value before conversion (upper limit value of reference range for inspection value before conversion)
RM: Median of reference range before conversion (median of inspection value before conversion)
oRL: Lower limit value of target reference range (lower limit value of target test value reference range)
oRH: Target standard range upper limit value (Target test value standard range upper limit value)
Here, when it is determined that even one of the above parameters does not exist, the non-parametric conversion process is terminated because conversion is impossible.

Next, based on the reference range lower limit value RL, upper limit value RH, and target reference range lower limit value oRL, upper limit value oRH, the range TR of the test value before conversion and the standard of the target test value are calculated using the following formulas: The range width oTR is calculated (S32, 33).
TR = RH − RL
oTR = oRH − oRL

Next, based on the reference range lower limit value RL before conversion, the upper limit value RH, and the median value RM, the estimated values of the upper hinge HH and the lower hinge HL in the distribution of the inspection value before conversion are calculated by the following formula (S34). .
HH = | RM − RH | × k (k = 25 / 47.5)
HL = | RM-RL | x k (k = 25 / 47.5)

Next, based on the widths TR, oTR, the upper hinge HH, and the lower hinge HL of each reference range, the estimated values of the upper hinge oHH and the lower hinge oHL in the distribution of the target inspection value are calculated by the following formula (S35). )
oHL = HL / TR x oTR
oHH = HH / TR x oTR

Next, based on the target reference range lower limit value oRL and the lower hinge oHL in the distribution of measured test values, the estimated median value oRM of the target test values is calculated by the following formula (S36). In the following formula, k is a coefficient (k = 25 / 47.5).
oRM = oRL + (oHL / k)

  Next, it is determined whether or not the pre-conversion inspection value Xi is greater than or equal to the median value RM (S37).

If it is determined in step S37 that the pre-conversion inspection value Xi is equal to or greater than the median value RM, the standard hinge value HXi is calculated by the following equation (S38), and the post-conversion inspection is performed based on the calculated hinge value HXi. The value aXi is calculated (S39), and the nonparametric conversion process is terminated.
HXi = (Xi−RM) / HH
aXi ＝ oRM ＋ HXi × oHH

On the other hand, if it is determined in step S37 that the pre-conversion inspection value Xi is not equal to or greater than the median value RM, the standard hinge value HXi is calculated by the following equation (S40), and the post-conversion inspection is performed based on the calculated hinge value HXi. The value aXi is calculated (S41), and the nonparametric conversion process is terminated.
HXi ＝ (Xi−RM) / HL
aXi ＝ oRM−HXi × oHL

  Thus, according to the test value conversion program of this embodiment, the test value measured under one measurement condition can be compared with the target test value measured under another target measurement condition only with easily available parameters. Can be easily converted into a simple value. In addition, the post-conversion test value calculated by the test value conversion program of the present embodiment is converted based on the distribution type of the test value. Therefore, even if the number of parameters is small, it is comparable to the conversion value by the correlation method. A highly reliable inspection value after conversion can be obtained.

  The inspection value conversion program of the present invention is not limited to the embodiment described above, and various modifications can be made without departing from the spirit of the present invention.

  For example, the test value conversion program of the above embodiment is a program that collectively converts the test result data sent from the test facility, but the test value conversion program of the present invention individually converts the test result of the individual subject. It may be a program for conversion.

  In the test value conversion program of the above embodiment, the distribution type list for the attribute of the test value is fetched from the outside, but the test values of healthy persons measured under the same measurement conditions are collected. The list may be updated by determining the distribution type from the distribution of the inspection values.

  Further, in the inspection value conversion program of the above embodiment, the reference range of the target inspection value is listed in advance, but the reference range of the target inspection value is set every time the inspection result data is converted. Also good.

  Moreover, although the test value conversion program of the said Example utilized the reference | standard range and median value of the test value before conversion contained in test result data, distribution of the test value before conversion of the healthy person contained in test result data From this, the reference range or median value of the inspection value before conversion may be calculated.

  In the inspection value conversion program of the above embodiment, the distribution type and reference range of the inspection value are determined by the inspection item and the gender of the subject. It may be determined by further subdividing according to the age group or nationality of the examiner.

  In the inspection value conversion program of the above embodiment, the inspection value conversion processing is not subject to conversion for inspection values for which only one of the upper limit value or the lower limit value of the reference range is set. However, such inspection values are converted separately. Processing may be performed.

  Moreover, in the inspection value conversion program of the above embodiment, all inspection values before conversion are converted based on the distribution type. However, the inspection value conversion program of the present invention uses the conversion and distribution by the correlation method. You may use conversion based on a type | mold together. Specifically, the inspection item for which the conversion method by the correlation method is established may be converted by the correlation method, and the remaining inspection items may be converted based on the distribution type.

It is explanatory drawing which shows the outline | summary of a normal distribution type conversion process. It is explanatory drawing which shows the outline | summary of a non-normal distribution type conversion process. It is explanatory drawing which shows the outline | summary of the non-normal distribution type conversion process continued from FIG. It is explanatory drawing which shows the outline | summary of a nonparametric conversion process. It is a conceptual diagram which shows the outline | summary of a test value conversion system. It is a flowchart which shows the control processing content of a test value conversion process. It is a flowchart which shows the control processing content of a normal distribution type conversion process. It is a flowchart which shows the control processing content of a non-normal distribution type conversion process. It is a flowchart which shows the control processing content of a nonparametric conversion process.

Claims (4)

A program for causing a computer to execute a process of converting a pre-conversion inspection value measured under one measurement condition into a post-conversion inspection value that can be compared with a target inspection value measured under a target measurement condition different from the measurement condition. ,
The same deviation value as the deviation value of the pre-conversion test value in the distribution of the target test value based on the pre-conversion test value having a normal distribution type, the reference range of the test value before conversion, and the reference range of the target test value A test value conversion program for health examination, which causes a computer to execute a normal distribution type conversion process for calculating a numerical value to be a converted test value. A program for causing a computer to execute a process of converting a pre-conversion inspection value measured under one measurement condition into a post-conversion inspection value that can be compared with a target inspection value measured under a target measurement condition different from the measurement condition. ,
Converting the non-normal distribution type pre-conversion test value, the pre-conversion test value reference range, and the target test value reference range into variables so that the non-normal distribution type distribution is a normal distribution type, respectively,
Based on the test value before conversion after variable conversion, the standard range of test value before conversion after variable conversion, and the standard range after variable conversion of target test value,
In the distribution obtained by converting the target inspection value into the variable, a numerical value that is the same deviation value as the deviation value of the inspection value before conversion after the variable conversion is calculated,
Further, a test value conversion program for health check, which causes a computer to execute non-normal distribution type conversion processing in which the numerical value is subjected to inverse conversion of the variable conversion and calculated as a converted test value. A program for causing a computer to execute a process of converting a pre-conversion inspection value measured under one measurement condition into a post-conversion inspection value that can be compared with a target inspection value measured under a target measurement condition different from the measurement condition. ,
Calculate the estimated median in the reference range of the target test value to be the same ratio as the ratio of the median in the reference range of the test value before conversion,
Subsequently, the inspection value before conversion is compared with the median of the inspection value before conversion,
When the inspection value before conversion is larger than the median value, the ratio of the difference between the inspection value after conversion and the estimated median value relative to the difference between the upper limit value of the reference range of the target inspection value and the estimated median value is the inspection value before conversion. Calculate the inspection value after conversion that is the same as the ratio of the difference between the inspection value before conversion and the median value relative to the difference between the upper limit value and the median value of the reference range,
If the test value before conversion is smaller than the median value, the ratio of the difference between the estimated median value of the target test value and the lower limit value of the reference range to the difference between the estimated median value and the test value after conversion is the center of the test value before conversion. Non-parametric conversion processing for calculating a post-conversion test value that is the same as the ratio of the difference between the median value and the pre-conversion test value relative to the difference between the value and the upper limit value of the reference range. Conversion program. Conversion in which a pre-conversion test value measured under one measurement condition can be compared with a target test value measured under a target measurement condition different from the measurement condition in a computer storing a distribution type list for the attribute of the test value A program for executing a process for converting to a post-inspection value,
From the attribute information of the pre-conversion inspection value acquired in advance, a distribution type determination process for determining the distribution type of the pre-conversion inspection value based on the distribution type list is performed,
If it is determined in the distribution type determination process that the distribution type of the inspection value before conversion is a normal distribution type, the computer executes the normal distribution type conversion process according to claim 1,
If the distribution type determination process determines that the distribution type is a predetermined distribution type other than the normal distribution type, the computer executes the non-normal distribution type conversion process according to claim 2,
4. A test value for health check, which causes a computer to execute the non-parametric conversion process according to claim 3 when the distribution type determination process determines that the distribution type is neither the normal distribution type nor the predetermined distribution type. Conversion program.

Images