CA2050320A1 - Medical statistical analyzing method - Google Patents

Abstract

A method for deriving statistical information and control charts in a patient. The process comprising steps of: (1) selecting data from a monitoring device, (2) forming data into records, (3) selecting a portion of each record, (4) determining the average-average, and the standard deviation, (5) repeating steps (1)-(4) after a delay, (6) setting control limits, (7) set-up control charts on a moving screen, (8) determining average-average and standard deviation, (9) graphing and displaying sigma, x-bar and range against the control charts, (10) charting statistically significant deviations, (11) re-adjusting the x-double bar and sigma and/or the R chart sigma bar and control limits, (12) repeating steps (5)-(12), (13) sending data from control chart to a database, (14) comparing isolated portions to database segments, (15) matching categorized portions to treatments and diagnosis, (16) grouping sets of diagnosis/treatments, (19) displaying on the screen options of diagnosis/treatment.

Description

2~32~
WOg~ 395 / PCT/US91/0~2 ~DICAL STATISTICAL ANALYZING ~HOD
BACX~Qy~ o~ IMvÆ~Q~
2~ Prio~ Ar~
This invention applies to statistical devices.
More particularly, this invention applies to me~ical statistical devices or devices using statistics in a medical environment.
More particularly, the invention applies to obtaining statiBtical information in control charts, for monitoring and analysis of medical processes.
Statistical analysis using Shehwart type process control charts was developed approximately 40 years ago in its present form. The use of statistics generally in medicine probably dates back to the early use o~ the scientific method in determining causes. 5tatistics are currently kept ~or pu~poses oP di~ea~e control and diagnosis. The ~hortcoming of prior art in thiis area is not the failure to use accepted statistical techniques, ?
but the failure to treat a medical patient as a true process. In any process, it is unacceptable to have an inflexible standard set because of the number of changes and steady states possible in any situation.
One example of statistical analysis in a medical environment is found in Bell, et al.; U.S. Patent No.

3,322,954; which relates to diagnosis of statistically significant varia~ions of radiation.
The use of inflexible standards in a statistical situation using central processing units is also known in ~he medical field. Hutchins; U.S. Patent 4,583,524;

~. : . ;, ; . . . : .. . . , . . .... ,. ; . .~ ~ , v W091/10395 ~ PCT/U~91tO02 ~ ' processes inf ormation in order to obtain medical diagnosis or treatment. Hrushesky; U.S. Patent

4,5l9,395: shows the use of statistically analyzed mean and standard error in heart rate. It ~oresees the monitoring of not only the patient, but the addition of drugs to the patient over time. Similarly, John; U.S.
Patent 4,545,388; shows the application of basic statistical computation o~ mean and variance and comparison to a previously obtained self norm ~or a given individual. Both of these patents substitute the use of statistically inflexible standards ~or medical judgment and provide signals relative to the change. Both ~all short of ~he current invention by ~ailing to provide ~or proces~ charts and ~ollowing the various stable states in the course o~ a patient's tr~atmen~ ~he~e patents substitute limited information to the information provided by Shehwart type process control charts which signify change over time and provide a continuous and monitorable statistical analysis.
Hrushesky and John address the broad aspects applicable to basic determinations o~ the idea of finding statistically significant changes in mean and standard deviation or variance measurements. Because neither use control chart tracking of a patient, the prior art patents are li~ited to situations where a known norm is available and where the patient is controlled only by attempting to reach the given norm.
The pres~nt invention allows for the variations .

20~3~
WO9~/10395 3 PCT/US91/00254 necessary to follow a patient who does not have a norm, and perhaps never will, during the medical intervention process. The present met~od treats the patient as an ongoing process with the examination being directed to changes in the process and stabilizing or maneuvering the process in any given direction. The use of control charts allows the physician to set a given norm for a patient regardless o~ the patient's current condition or records kept on the patient.
The present invention addresses the question by giving a graphical analysis which is continuously monitorable by the physician and wherein the limits may be adjusted to allow the user to reset the analy~is as a given patient ehanges.
Typical medieal deviees u~ing eurrent keehnology only give indieato~s or alarms o~ problems w~ieh show single events out o~ the ordinary when a patient's condition has already become unstable. Prior art was designed in ord~r to have machines assist in the praetice medicine for the doctor. Sinee medicine is a less than certain science, this results in equipment which does not serve a consistently useful function statistically. -~ -Other existing equipment and methods provide a graph format but without statistical ~nalysis, merely provide graphing of raw data. The present process allows for obtaining statistically significant historical analysis of the varying conditions, medicine, and equipm~nt used ~or treatment.

` . ` .... . , ., . , . . ' .. .` .. " :, . . . ~. .... . .; : ": 1. .. . ` . . . .. ` .

WO91/1~3~5 2 ~ ~ 0 3 ~ ~ P~T/VS91~002 ~
One purpose of this invention is to provide a physician statistical information on a patient and to present the information in an interpretable form to a physician~statistician at a ~onstant rate with sufficient statistical information being provided at one time to have statistical significance.`
Another purpose of the method is to provide an early warning system ~or medical patients. This process will e~hance the early recognitions o~ problems with m~lical patients while being monitored with various equipment.
Another purpose is to provide an early warning system which will enable the physician to reduce the probability of the patient's going into ~nstable conditions. This process will also enable physicians to determine the stability of a patient ~or ~he proc~s~ o~
discharging from hospital skay.
Another purpose it to provide a process with the capability to inter~ace computers to medical devices.
The process will use serial port communications from medical devices to a computer. The computer will then statistically analyze data received from t~e ~edical device and graphically display the statistical analysis of this data. This analysis of ~ata will enable the physician or clinician to detect early indicators of non~stable conditions for the patient.
Another purpose is to provide for alarm systems based on an unknown patient; they are not individualized.
The statiætical analysis and graphics are based on an .

.. .. ...

~ Wog~/1039S 5 2 ~ ~ ~ 3 2 ~ PCT/US91/00254 established normal for the individual. Mathematical laws and statistical formulas establish cc~.trol limits for the patient based on the normal values for a particular patient.
Another purpose is to provide a method o~
statistical analysis which will provide statistically significant in~ormation about medical equipment and medications applied during the monitoring period.
These and other obje~ts and ad~antages o~ the invention will become better understood hereinafter from a consideration of the specification with refcrence to the accompanying drawings ~orming part thereof, and in which like numerals correspond to parts throughout the sev~ral views of the invention.

The system includes hardware including patient interfaces with medical devices, communication cables to medical devices, the medical device itself, communications devices from the medical devices to computers, the computer e~uipment which receives the cable ~rom the medical device, a separate input to record on a time coordinated basis treatment into the computer from an operator or fro~ automatic treatment e~uipment.
The equipment also utilizes so~tware which includes a data communication software, storage software, statistical analysis software, di~play so~tware for process contral charts, ti~e dating so~twar~, and '' `, ;' 2 ~ ~'3 3~ PCTtU~91/~02 software for communication with the host database and for getting replies from the host database where tha host database comprises data ~hich has been previously analyzed stati~tically re~ative to a large number of patients.
~ rior art which simply reads if a continuous number of elements fall outside the control chart fail to give an adequate amount o~ information to som~one using this for statistical analysis purposes. In this invention, the use of the theory of runs to analyze where a significant number o~ deviations fall on a given side of the average for a control chart. The number which is statistically significant is set in step 6.
The software i5 an integral part oP a network which acts on readin~s from a pati~nt in order to produce a readable analysis of the patient's condition~
Medical devices using current technology only give indicators or alarms of problems when a pa~ient's condition has already become unstable. An example of this is a normal alarm for patient heart rates, or a rate of 40, which is an indicator of bradycardia of a patient, or a rate of 140, on the high end, which is an indicator of tachycardia of a patient. When these alarms are ~iolated, the patient is already unstable, resulting in treatment by the use of drugs or the use o~ medical devices. It has been documented in medical journals that the treatment with medical devices such as a defibrillator to return a patient to stability causes , ,:.. , . , ~ . ~ . . , ,, :......... .. . ..

~ WO91/1039~ 2 ~ ~ ~ 3 2 ~ PCT/US91/0~2~4 damage to the heart muscles. This invention uses statistical analysis o~ heart rate that a physician would be able to reco~nize as an indicator of an unstable condition earlier than he would with normal alarms. This indicator would allow the physician to treat the patIents with drugs or medication and nok more extreme medical devices such as a.defibrillator. This would result in less long~term damage to that patient's heart.
Another example is the use of a patient's temperature. The indicators of a patient's temperature are typically not monitored by an alarm, and only viewed randomly by the clinician~. An increase in a patient's temperature is an indicator of a patiént having an infection. The use o~ graphics display and ~tat~stical analysis o~ tempera~ure can re~ult in an early warn~ng o~
a potential in~ection. This would allow the doctor to possibly totally eliminate the infection or reduce the effect on the body of an infection.
Another application is interfacing to pulse oxymetry. Pulse oxymetry is used to measure the oxygen concentration in the bloodstream of a patient. Normal alarm conditions are preset at a rate of 90~ oxy~en saturation. Once the patient has reached a level of 90~
oxygen sa~uration, there is an immediate need for oxygen to be administered to the patient or permanent damage to the patient will occur. Through the use of our statistical analysis and qraphics displa~, an early indicator o~ reduction in oxy~en satura~ion will occur, , '; ..... ' '' ' ? ,, WOgl/l0395 2 ~ ~ ~ 3 2 ~ PCT/US91/0025 ~
thus allowing the physician to utilize oxygen treatment prior to the patient going into duress and having possible da~age.
Another application ls for the monitoring o~
invasive blo~d pressure during surgical procedure.
Invasive blood pressure during surgical procedures are normally monitored with no alarms and are observed for radical changes by the anesthesiologist in the case.
Typically, when a large change occurs, the anesthesiologist will be forced to administer high amounts of drugs or make a large change in the an~sthetizing agent which is used to keep the patient unconscious during the procedure. ~hrough th~ use o~ our statistical analy~is and graphic d~splay, the anesthe~iologist is allowed to recognize th~ changes in the patient earlier. This early recognition allows the anesthesiologist to cbntrol the patient with a smaller dosage o~ medication or a lesser percentage of anesthetizing agent~ It is documented in ~any medical journals that long-term health is increased with a reduced amount of medications or a reduced amount of anesthetizing agents.
Another application of the statistical analysis is in the use o~ la~oratory results on the patient, such as Ph level ~r oxygen level~ Typically, laboratory results are looked to be within a very wide window and are treated to be either acceptable or unacceptable. By the use of the statistical analysis and graphic display in :~ , .,, ., ....... . , , ~ ... .,.:

2~503~
~ W091/10395 ' PCT/US91~002~
.~ ~ 9 this method, as a change occurs in a particular patient that is outside of 3 standard deviations for that particular pat~ent, an early indicator of potential problems with the patient is announced.
Examples of specific types of equipment that can be interfaced are:
1. Patient physiological monitors, which include the monitoring of heart rate, pulse rate, respiration, non-invasive blood pressure, invasive blood pressure, and temperature.
2. Pulse oxymetry, which monitors oxygen saturation (S~O,), and pulse rate.
3. Non-invasive blood pressure apparatus 4. Devices ~or the measuring of oxygon.

5. Devic0s Por the mea~uring of carbon dioxide. ;~

6. Capnograms.

7. Devices for the measurement of entitled carbon dioxide.

8. Devices which electronically measure urine output. ;

9. Devic~:s which measure temperature. `~;
,:

.

..... ~ .. - . . . . , .. , . ,., . . . .. , . ~.. ; . . . .

,, . ; '. ` ' . . . ' ' , : ., ,, , : , , ' , , ! :
' '" ` ' "`.. ' . "'''' ''`' '.. , . '' '. ' .. : .. .. .. .

WV91/1039~ ~0~ 0 3 2 ~ PCT/US~ 02 BRIEF ~ESC~IPTI~b~5~3l~ o I~5 For a furthar understanding of the nature and objects of the present invention, re~erence should ~e made to the following detailed description taken in conjunction with the accompanying drawings in which like parts are given like reference numerals and wherein:
Table l is a block diagram o~ the process steps ~or setting up hardware utilized in collecting data for the invention.
Table 2 is a block diagram of the ~rocess steps used for setting up a system used by the invention.
Table 3 is a block diagram of the process used for collecting data and establishing control charts used by the inventi~n.
Figuxe l is a plan view o~ a patient monitored by the the invention.
Figure 2 is a graphical representation of a set of three charts generated by the system showing the effect of a change of average without a change of standard deviation.
Figure 3 is a graphical representatlon of a set of three charts generated hy the system showing the effect of a change of standard deviation wi~hout a change of average.
Figure 4 is a graphical representation o~ a set of , .
three charts generated by the system showing the effect o~ a change of standard deviation with a change o~
average. ~.

... ~ :: . ::: . :: . .. .. .

~ W~9lJlV395 2 ~ ~ ~ 3 2 ~3 PCT/US~lJ0025~
Figure 5 is a graphical representation of a set of three charts generated by the system showing the effect of a change of standard de~iation with an increasing change of average.
Figure 6 is a graphical representation of a set of three charts generated by the system showing the e~fect of a change of standard deviation with a decreasing change of av~rage.
Figure 7 is a graphical representation of a set of three charts generated by the system showing the effect of a change of average without a change of standard deviation.
Figure 8 is a graphical representation o~ a set o~
three charts generated by the ~stam showing tha e~fec~
of a change o~ standard deviation without a chanqe of average.
Figure 9 is a graphical representation of a se~ of three charts generated by the system showing the effect of a change of average without a change of standard deviation.

.

~; , , , ,, ~ , , ! . , WO 91/10395 2 0 ~ ~ 3 ~ ~3 PCr/US91/~025~i DET~IL~2 DIS~USS~ON of the PREFEF~Q~TL~) The purpose o~ the invention is in order to allow for the application of quality control concepts to existing equipment used in a medical environment.
Typically, the equipment which exists in a ~edical environment produces data which is in digltal ~orm in the form o~ ASCII, or else is an analog ~orm which can be translaked to digital by any number of means which are known in the art of analog to digital conversion.
For purposes of the discussion, it Is assumed that the signals are digital si~nals and that the conversion step is unnecessary, although the use o~ the ~nvention wit~ an analog signal to a dlgital ~ignal ~or purpose~ o~
this would be identical except ~or the added step of con~ersion.
In order to avotd confusion, numbers rePerencing drawings appear in numeric form and numbers otherwise used in describing the invention are typed out alphabetically.
Communications ~etween computer equipment usually require at least three pins, but m~y use twenty-five-pin standard communication port which is present on most modern medical analysis or reading e~uipment.
For purposes of this invention, the data which is received is sent through pins which are provided on the equipmant to an outside source, ~ut the same technology would apply if the data was used internally in ~the .. .... -.: .. .. . . . ... ... . .. .. ...

2Q~3~
WO91/10395 PCT/US9~/002 machine or medical and the medical equipment was supplied with screens amd the other components necessary in order to ef.ectuate the invantion It is noted that any slngle piece of me~ical equipment may send out several different medical bits of information, for example, blood pressure readings and temperature, and in addltion may have control characters added to the readout ~rom the exlsting equipment.
The first step of the process, there~ore, is to take the data which is coming in blocks which can include control characters, temperature and blood pressure and . :
separate out the specific data to be used utilizing methods known in the art from the blocks of data.
Table l:Hardware setup ;.

(25) Establlshing Communications between a medlcal instrument and a microcomputer using techniques known in the art (26) Determining the mëdical data to be ~onitored,¦
stored, analyzed, displayed as controlled charts.
~his is done by selection by a medical doctor of ¦
what information should be monitored in a patient. ¦

~(27) Determining the medical data to be monitored,~
stored, analyzed and displayed ~28) Creating a procedure to identi~y, isolate and : ~ capt~re the ralevan~ data within each data record.

Referring to Figure 1, the entire process can be described in. ~teps, with the first step l being the selection o~ standards. ~hese standard5 may be from a , : .:: , ~ ~ ' . `` - .

Wogl~lo39~ 2a~32~ PCT/US9~/0~2 ~
preset group.
Other information may he desired which could include selecting the statistical method to be used. The types of methods available, generally under this invention which uses control charts would be ~rom the following sets:
(a) Average and sigma charts Average and range charts P charts, sèttiny out the percent effective or de~ective U charts or > defects or defects per instance C charts NP charts - number of occurrences In the prQferred embodiment, the average and sigma charts only are used. The use range charts is also set out parenthetically as an alternative t~ the ~igma chart.
It is to be no~ed that stati~tically the ran~ chart i~
only an approximation and in the pre~erred embodiment would not be used.
In utilizing the invention, the ~irst step is the set up of ha~dware. Referring generally to Table 1 this is accomplished by establishing step 25 between hardware shown in Table 4 as the medical instrument 20 and microprocessor 21 and database 23. This assumes the existence of communications be~ween patient 24, th~ough medical device interface 22 and medical instrument 20.
Referring again to Figure 1, determining (by mental operation of the user) the medical data to be monitored, stored, analyzed and displayad as control charts is determined step 26 ~rom the d~ta coming a~ digital .. ~' ,.

20~ 2~
W09l/10395 P~T/US91/0~2~4 signals output by the specific monitoring equipment. The determlnation is a selection by the use.- ~f the specific con~itiGn (e.g temperature, ~loo~ pressure, etc.) to be monitored using the invention. Next, determining the data record ~rom the data coming into the system is isolated step ~7 using data block separation techniques known in the art. Finally, in the hardware set up, creating the procedure to identify, isolate and capture thd relevant data within each record from known techniques for working with computer signals is step 28. -;
This type of hardware set up is commonly used by existing equipment. One di~ference in the present invention from the existing equipment is merely the isolation of the in~ormation ~or modi~ication in the proce-qs ~t~ps which follow.
Table 2: System set up .... _ .__ (1) Determining the sample subgroup (n) usually 5, often in the range 1-9.
, . .. 1. . __ ... ~ . __.
. . _ I
(2) Determining the method for introducing a delay between subgroup sampling.
...... __ I
.... _ ~
(3~ Determining the number of subgroups (k) in the initial sample (usually 15-25) i I
I
(4) Determining the statistical method to be used.
Common selectic~s include:
Average and sic-:~a charts Average and range charts P chzrts U charts C charts .,. ~ . .

W~91/103~ 2 0 ~ ~ 3 2 ~ PCT~US91/002 ~

. . ~
(5) Determining confidence level uslng probability or nu~ber ~f standard deviations. Choice includes:
2 sigma warning limits and 3 sigma actions limits 4 sigma + if there is a problem with co~variance that cannot be avoided due to medical or other reasons.
I .
(6) Determiining the theory of runs ( e.g. 8 in row, in row, 4 of 5, 2 of 3, and 1 outside).
. . _ ___ I . . . ~ , ..
(7)¦SpeciPy what data is to be s~ored.
., . . _ .._ I
(8) Specify what data is sent to a database.
a. Size b. How Data Collection and Establishing Control Chart (9) Input or Collect a data string a. Dividing data string into records.
b. Stripping d~ta Prom record~ with minimum delay betwean each .
_ _ . . . ................. _ (lO) Selecting a siubgroup o~ size n ~rom the data selected.
. .... ~
11.0~ tll.l) I
Analys:.s: Medically Specified:
a. Calculate average a. Average b. Calculate sigma(or range) b. sigma or range c. storing the calculated average and sigma I I
(12)¦Repeat steps 9-11 k times after time delay between each I
. I
(13) Calculate values for control charts a.Calculate average of averages ~x double har) b.Calculate average of sigma (sigma bar or r bar) - i, . ; . . ~ ; . . . .. . . . . . . ..

03~
W~91/10395 PCT/US91t00254 (14) Calculate control limits ~or a. For average control chart b. For si~m~ control chart ~or range) .... _ ~.
t15) Generate contrQl charts a. Averages b. Sigma or range . . ~ . . . ~
(16) Instrumentation and non-medical causes a. Identi~y all possible instrumentation and non-medical outliers.
b. Ellminate the causes, i~ possible.
**~xperience has shown that it i~ usually necessary to ~ake a t~lal run for each type of instrument used before applying the technique to an actual patient in order to eliminate non-medical causes for outliers. Once a system has been developed for a particular characteristic this step is usually not required.
I , ~
, (17) Establish control charts on a change cau~e system ~ree o~ as many non-modical Gau~es o~ variation.
-Run control chart~ in ~oreground or background.
a. Index moving bar to next po~tion~
b. Calculate pos~tion o~ average.
c. Calculate position o~ sigma.
d. Identi~y and mark statistically outliers (outside limlts and runs).
(17.1)~ (17.2) I ~
Graph and display. ¦Display results (out or in)¦
I
. _ - . :~ ' (18) a. Isolate data specified.
b. Send data to a database.
c. Medical database analysis and matching.
d. Comparing the isolated portion to the database collection of similar seyments and categorizing the same.~
e. Comparing said portions to accepted treatments and diagnosis.
f. Grouping such sets of treatmants and dlagnosis w1th the corresponding portions and ~5U~
WO91/1039~ PCT/US91/002 .~ .
~l9) a. Isolate data specified.
b. Send data to a database.
c. Statistical database analysis and matching-pattern recognition.
d. Comparing the isolated portion to the database collection o~ similar segments and categorizing the same.
e. Display treatment and diagnosis with the corresponding portions and f. Initiate treatment Subsequ~nt to hardware setup, the system must be set up as shown in Table 2. Standa~ds are determined set up.
Determining the standa~ds include the sample subgroup step l, (~eferred to as N in algorithms used herein); (23 determining the method for introducing delay step 2 `
between subyroup samples step 1 by programming techniques or by having delays built into the computer equipment analyzing the data ~tream ~rom t~e monitoring e~uipment based on the need ~or statistlaall~ silgni~ican~ delays between samples a~ known in the art; the number or set of subgroups or repetitions of subgroups in the initial sample step 3 (Referred to as K in algorithm~ used herein).
The set step 3 of subgroups step l are used in the preferred embodiment for determining the average average (x-double ~ar) and average standard deviation (sigma bar).
Also in the setup, the user must determine the statistical method to be used 4. As indicated, t~e statistical method to be used is usually average and sigma charts (other options would include average and '.

20~ 032~
WO91~10395 PCT/US91/002~4 range charts, p charts, u charts or c charts).
Also in ~he setup, the use must determine the confidence level step 5 for the control charts using probability or the number of standard deviations.
Choices include a factor times the average standard deviatio~ (sigma), the theory of runs as descrlbed in more detail below, or similar methods known in the art of statistics.
Standard deviation choices such as TWO sigma warning l1mits (i.e. marking occurrences which are outside TWO
sigma warning limits) and T~EE sigma action limits (taking appropriate action for occurrences outside Three Sigma) are examples. Four sigma or higher limits may be used where there are problem~ with co~variance that cannot be avoided due to medical or other reasons.
The theory of runs is determined step 6 and refers to the number of consecutive points graphed on either side of the average either with or without at least a certain number of the points being outside the control limits as set by the confidence level step 5.
In the preferred embodiment, the confidence level or ~ -sigma factor step 5 is the same for the upper and lower control limits and for both charts. Varying these levels step 5 so that the upper and lower control limits were different would not materially depart from the inventive con~ept used herein.
The next step in the system set up i~ determining the information to be stored step 7 and the information .

. :.. : : . , . :, , ~ . . . . . . . . . .

.. . . . . . ~ . .. . ~ .

W~ 91/~0395 2 0 ~ ~ 3 2 0 PCT/US91/002 ~
~D
to be sent to a database 23. An example of step 7 would be the selection of the size of the data sample to be sent step 7(a) and whe~her the sample is automatically sent or to ~e sent manually step 7(b). In the preferred embodiment the data to be stored step 7 is all of the data and the size of the data to be sent step 7(a) is the 30 points displayed and selection of a data sample step 7~b) is done manually by a particular keystroke on the equipment.
All of the settings set out as in steps 1 through step 7, or any group thereo~, could be pre-set into the equipment as a standard for all patients, entered individually, or keyed into the equipment automatically upon determining a giv~n con~ition o~ the patient ~rom a selection o~ conditions on the ~creen or may be kayed in one at a time.
Table 3 outlines data collectlon and establishing charts. The first step is the input of data step 9.
This may be done by hand, for example putting readings from notes or equipment on a time related basis into the system. By handling the input o~ data in this fashion, the need to s~parate computer signals, by various methods known to the art, steps 27 and 28 is avoided.
In the preferred emkodiment, this second step 9 assumes that the data will come about normally in a single block of data, which may include multiple readings. F~r example, a single temperature reading, blood pressure reading, and other medical readings, along ... .. ..... :.. :,.,.,. .. , .. , ...... , ,. , , ,... : . . .
.: .:: :, , ,. . . .; :.. . . . .. . : ~ ., ~ .: . .... :. . ..

~ WO91/10395 2 ~ ~ 0 3 2 ~ PCT/VS91/0~254 with a single set of control characters which the monitoring equipment generates for programming reasons internal ta the e~uipment may all co~e in a block ~rom which th~ specific data to be graphed is isolated by the techniques set up in steps 27 and 28.
Inputting data step 9, can be broken into steps ~(a) dividing o~ data into records, stripping data step 9(b) which is the isolation of the datum to be processed, which, for purposes of this discussion, will be assumed to be the temperature, and setting aside the other portions of a given block which could be treated in a similar fashion, except the control characters probably would not be utilized~
Select-ing ~r placing data ~t~p lO of ona o~ a ~t the number of whiah is specif~ed in step 3 of a certain number tk) of consecutive subgroups the size of which is s~ecified in step l of size n, allows for putting togetner subgroups which are preferably sets l of four or more stripped data units, but sets of at least one stripped data units as se~ in step l.
The number, k, of such subgroups step 1 could be reduced to one ~or extremely slow readings; two would yive a distribution, but a better distribution for purposes of statistical analysis is derived from subgroups step l which have ~or n at least four to seven individual readings set in step I or datum. The group~
or sets specified in step 3 of subgroups l ne~d only approach being consecutive and the failure to have the w091J10395 2 0 ~ 0 3 2 0 PCT/US91/002 ~
groups input be perfectly consacutive would not materially depart from the inventive concept herein.
An example of steps 9 through lO would be as follows: Each data point or reading wou~d be isolated in step s(a) and 9(b) and grouped in subgroups of a size specified in step l of size n. In the preferred embodiment n would be equal to step 5. This subgroup would then be selected step lO by the user or by the program and the analysis which follows would take place.
For each subgroup selected in ~tep lO one of two steps would be availa~le. Analysis step ll.0 or Medical specification step ll.l for average step llta) and sigma (or range) step ll(b). ~ Medical specificatiQn is used, steps 12 th~ough l~ may be skipped ~or purposes o~
generating control charts s~ep l5 a~ 5et eorth below~
Analysis step ll.0 o~ the data in the subgroup step l includes step ll(a) calculating an average (x-bar) and step ll(b) calculating the standard deviation (sigma).
Calculating the range (R) of the set of the subgroùp step l for purposes of graphing at a later time is an alternative to calculating sig~a. Range is the statistical term for the dif~erence between the highest reading in the subgroup from the lowest reading in the subgroup. Range is used to approximate sigma.
Typically, the range ïs less accurate than sigma and therefore not desired.
This analysis step ll~0 would take place after each of the su~groups of step l sQlected in step lO are .. ' . ~ ~.;i ' ' ' ' " ' ' ' ~' ' '' ' ' ~ W~ 91tlO395 2 ~ 5 ~ ~ 2 ~ P~TtUS91/~0254 collected, utilizing the process set fort~ above, keeping the units as close together as conveniently possible in order to have ~ore cr less contin~ous samples in a subgroup of step 1.
This information is stored ll(c), the average and standard deviation (or range) ~or purposes of graphing at a later time on a bar chart AS described later.
Step 12 is a reptition of steps 9-11.0, usually with a delay set in step 2 ~or the sets step 3. The delay set in step 2 is usually accomplished by the programming but which may be a factor built into the sampling sy~tem due to delays in the computer equ.ipment analyzing the data stream ~rom the monitoring equipment.
Having the sets set in step 3 too close togethQr ~ould result in covariance with one reading affectlng the next and this would mean the control limits would be too tight and would not give the body or the process being analyzed a chance to c~ange. The delay 2 is a statistically signifi~ant delay 2 which would vary with the particular type of body reading, but is typically a very short time counted in seconds or portions af seconds.
The process above is repeated step 12 as set out until a statistically significant number of subgroups 1 are obtained. This number of subgroups is a set of a size speci~ied in step 3 of size k. The true statistical number necessary would ultimately be obtained ~rom obser~ations over a number of patients. In the preferred embodiment set o~ step 3, k, is nine to twenty-five ,:. . . : , . . .. ~- :
:.. : . . ....... . . .. . ~ .

20~3~i~091~03~ PCT~US9l/002 ~S~
subgroups of a size specified in step 1 of size five.
Fifteen is the value used for k belowas an example.
Calculating control chart values step 13 would be the purpose of setting the limits for obtaining a set of ~tatistics charts known: as control ~harts. This is done for an average (x-bar) control chart step 13(a) and a sigma control chart step 13(b). A range chart may be substituted ~or the siyma control chart.
Calculating step 13 involves taking all of the averages which are then ~veraged step 13(a) to obtain the average average (x double bar) for the twenty five suhgroups. All of the set sigmas step ll(b), are averaged in order to get an average sigma (sigma bar) stop 13(b) for the ~et stcp 3 o~ subgroups stop 1 (alternativ~ to sigma av~rage, all o~ the rang~s step 13(c) may be averaged to get the average range (R-bar) ;
for the set step 3 of subgroups step 1.
The sigma bar is obtained, for example, by adding up each sigma obtained and dividing by the number k in the set of size defined in step 3. For purposes of the discussion, it is assumed that this statistically significant number of repetitions of size defined in step 3 which are selected for this process is ~iftean, and therefore sigmas one through fifteen are added up and then divided by fifteen in order to get sigma bar, or the average standard deviation. The average range and avera~e average are determined using the same p~ocess.
It aan be noted at this time that the two control f~ WO91/103~ 2 0 ~ ~ 3 2 ~ PCT/US~I/0~2~
~ ~i5 chart6 (e.g. sigma and average), though normally used together, are co~pletely independent and the use of one c~art without the other is ctatistically signif icantO
Figures 1 through 9 show the average (or x-bar chart~
displayed on a single screen below the sigma chart for three di~ferent sets of conditions.
Using the alternative step 11.1, the average-average, x-bar, and sigma are specified by the user, alleviating the need for steps 11 through 13 for generating c~ntrol cha~ts step 15. The purpose of medical specification step 11.1 is described in more detail below.
The method ~or determining the value o~ si~ma or sigm~-bar ~or use o~ settin~ control chart lim~t~ is known in the art. As an example o~ ~he method, the derivation of sigma for an "S" or sigma chart could be according to a formula:

~ - E2 (n~ 2n*c22] 1/2 C~/
~ :
where -o' = population standard deviation -a. = sample standard deviation n = p~pulation size c2 = 2/n * (n-2)/~(n-1)/2~
r = the gamma function, i.e. -O~ , I~,x) - ¦e t ~ dt, ~or x > O

,... . ., . . , ~ , . .. , ..... .. ... : ,.
:,; . , . ., , . ~ - ~ .. , : .

- . . ,. .. .. .: .. ,., .:., :. : . .. , , . , ., ~. .. . . . ., .. .,,: ~. .; .

', W~91/1~39~ 2 0 ~ ~ 3 2 ~ PCT/US91/00~ ~
An approximation is: :

a ;~ a The upper and lower control limits and center lines are calculated as follows for 3 sigma: ;

c~ r (~ ~ ) where -~rcL - ~ + 3 ~9 ;

LCL- a - 3 a~, Center line - aa a - average a As an example of the method, the derivation of si~ma for an "R" or range chart would be:

o" = R/d R - average range d~ - standard devia ti on UCL~ 3d3 d~ ~ 3cl2a~

: ... . .,,:: .:: ., .. . : . .. , : " , . , . , ., ~ : . .. , , , . .. . , . . ., . ::

20~3~
WO91/1039S PCT~USgl/0~254 LCL - 3d3~ 3d2a"

cen~er liIle - R

For X-bar charts, the control limits are calculated as follows:

UC~ - x ~ --LCL - x _ 3 a~

Center line - x x- mean a// - sta~dard d~via tior~

n - sample size.

These control limit sizes are known and tabled in the art. The factors to be used are loaded into the program as data to be used in the preferred embodiment as "B" conYersion factors. For ~xample, with sigma charts B ( 2 ) ~ C2 ~ 3 0~,/ a/ .
:
and B(1) - c2 - 3O~/o/

and the 'IB" factors are previously obtained using tables :: :. :: : . .::,.:,............ . , ,. ,. .. , : . .

W091/10395 2 0 ~ ~ 3 2 ~ PCT/USg~/002~
known in the art.
The control limits are set step 14 for the charts using the sigma bar obtained. Statistically significant control limits usually utilized in industrial processes are three-sigma, as set forth above. Sigma factors are similarly derived from the pre-existing art.
This user determination of sigma step 5(a) could be statistically as low as two ~or lower), but would typically not be lower than two, and could be varied as high as four (or higher), or it could be any fraction between two and four, depending on what analysis o~ the particular data over time yielded. For purposes o~ most industrial processes, and therefore used in this example and in the pre~erred embodiment, the deviation 5~a) i9 three-sigma. Again, ~he 5p~Ci~iC value oE the sigma factor used step 5~a) would depend on determinations which the user would make.
This process could be repeated for variable data, binomial data, and percentage data from the processes analyzed .
In the preferred embodiment, two charts are then generated step 15, the X-bar chart or average chart step 151a), and a sigma chart 15(b). Sigma charts and R-charts reflect the same information which is often not signi~icantly different. Range charts are only brie~ly discussed but could be substituted for sigma charts without departing from the inventive concept herein.
Each of the charts generate~ in stepis 15(a) and ~ WO91/10395 2 ~ 5i O ~ ~ ~ ` PcT/usgl/o~2s4 15(b) has an upper and lower control limit which was calculated step 1~ using the 5ig~a factors set by the user in step 5(a) set forth abo~e and a mi~dle line which is the sigma average (sigma bar) for the sigma chart step 15(b) and the range average (R-bar) for the range chart (alternate step 15(b)) and the average average (x-double bar) for the average chart 15(a).
An altèrnative step to step lS is also available and is a major innovation possible with the invention. This would encompass the concept of stabilizing the patient within a set range and deviation which the user would select. This would accomplished using medically speci-fied average and sigma values step 11.1. Although identical to ~e~s 15, ~tep 11.1 setting the average to be attained anc setting the standard deviation to be attained and setting up control charts as set out in step 15 would provide control chart limits to which the user desired to bring the patient. This is not a normal statistical application but is available where a differ-ent stable condition is desirable within known control limits and where the patient's reading~ can be carefully changed. This would be a non-dia~nostic use of the control charts and would be instead a method of treat-ment. Those charts ~tep lS could be used wlth the control charts.
The step 1~ is to generate contral charts steps 15(a) and 15(b). Although the preferred embodiment envlsions the use of visible charts, the method works .. :,: :. :.:, .. . : . .: ., . , :: : . ................ ... . ~. .

.: : :: : ., . , .: i. ~ . .. , . ,,, . :~ . : . ... : : .. :.: .. .

Wosl/lo39~ 2 0 ~ ~ 3 2 ~ PCT/USg1tOo2~

equally well if the charts are merely for purposes of analyzing data and are nev~r actually displayed. This ~ill be seen from the description which follows as to the use of the charts so generated.
Next it is necessary to identify step 16(a) and eliminate step 16(b) all possible instrumentation and non-medical outliers and eliminate the causes i~ possi ble. Expe~ience has shown that it is usually necessary to make a trial run for each type o~ instrument used before applying the technique to an actual patient in order to eliminate non-medical causes ~or outliers. Once a system has been developed for a particular cha~acteris-tic, this step is not reguired and is therefore not speci~ic to all claims regarding the inventien. This identifica~ion and elimination Gt~pS 16~a) and ~6tb) result in establishing control charts ~ree o~ the identified outli~rs.
Control charts are then established on a change cause system, display step 17 of the control charts, free of as many non-medical causes of variation as possible.
Control charts step 15 may be run in foreground (actual physical display on a screen or printed chart) or background (not physically displayed). Figures 1 through 9 show the results of foreground display of control charts of step 15.
I~ the charts are physically displayed in step 17, in the preferred embodiment lt may be by a continuous graphing using a moving bar graph o~ the type known in 20~32a /~v39s PCT/US91~00254 the art. At least 30 points are usually desired on the screen, although more points or less may be desired and the actual number o~ points on a particular screen would vary and would be a function of the relevance of the history and abilities of the screen. As an example, if there were 15 minutes of relevant history and a high sampling rate, then 30 points is usually su~icient.
Additionally, a script-type printout could be used in order to maintain a printout history.
Display step 17 requires that the data to be displayed be obtained. The data is obtained through the repetition of steps 9 through 11.0 ~ollowed by graphing the results agains~ the control charts. Steps 9 through 11.0 are repeated step 12 throughout the monitoring process in order to provide the in~ormation ne~ded ~or the display step 17.
The displaying step 17 may be broken down as intdexing in a time scale s` ap 17(a) a moving bar to the next position, 17(b) calculating the position of average on the x-bar chart, calculating the position of sigma on the ~igma chart, and identifying and marking statistical-ly significant outliers step 17(d) tpoints outside control limits or points outside the theory of runs).
This graphing step 17 is described in the discussion of Figures 1 through 9 below in more detail.
The sample rate may now ch llge within limits without a~ecting tne data which is produced. If the sample rate is too slow, possibly significant events could be missed :;: ~ . .. : . .- . i ,,. .: . . . . ~. ... . . .

wo gl/10395 2 0 ~ ~ 3 2 ~ PCT~US91/002 ~
and if the sample rate is too quick, the covariance problem resurfaces, as discussed above.
The speed with which data is displayed is such that the data should show the current condition of the patient, and that is what is strived for with any particular equipment.
The graphed samples of step 17 are obtained using the exact same steps set out in steps 9 through 11.0 in order to get the range, average and standard deviations of the subgroups.
Locating step 17(c) and marking step 17(d) points requires analysis. The analysi~ in locating step 17(c) is to analy~e the point to be plotted relati~e ~o the middle line o~ the chart. The analy~is in marklng step 17(d) requires determini~g i~ the point is outside oX the sigma limits set in step 5 using information calculated in steps 13 and 14 and whether the point violates the theory of runs 6. This is also discussed in the discus-sion of Figures 1 through 9.
~ aving at least two control charts, X-bar and sigma charts, i~ the preferred embodiment provides for two different methods of analysis for the information which is retrieved on a visual basis, which is not currently provided in medical technology. The first marking step 17(a) which is provided for in the two charts fo~ a point outside of the control limits is circling the variant data and an audible tone is glv~n may be in order to allow the particular measure~ent which fell outside of wogl/lo39s 2 ~ ~ 0 3 '3~ PCT~S91/~02S4 the control li~its to be noted.
An alternative would not display the actual graph (it woul~ be run in the ~ackground) but display the occurrence of statistically signi~icant indicators o~
change step 17.1(d). This could provide the same information without a constant display run in the ~oreground.
Because there is a continuous straam o~ this data~
if a pattern appears of points outside the chart, it can be recognized and, similarly, if only a single individu-al reading falls outæide, then it may be noted without having any reaction as a result of it.
Similarly, trends are shown by the X-bar chart as i~
moves in one direction, whereas destabilization~ become more apparent with the ~-chart or th~ sigma chart.
This method of interpreting patterns of variation on X-bar and R or sigma charts is documented in an industri al setting. Also, statistical analysis follows set patterns which these particular charts allow ~or the first time to be used in the medical field. However, the invention allows for the an~lysis and comparative use of information obtained from various patients over time, step 18, either as the treating physician noted a particular pattern which he wanted to review or on a continuous basis as spec~fied.
Formulation of a database step 18 comprises the steps of step 18(a) isolatinq segmentæ, as specified in step 7, of control charts, ~o~ example, a 30 item display ''','''., " ''.,, '' '', . ' ' ' ' ` '. ' `: ' ' ' ' ' `
' . ' ' :. '.,. . " ' . '. ' ' . . '.

WO 9~/lû3~ 2 ~ 3 ~ ~ PCr~US91/002~i of the patient, step 18(b~ sending this segment to a database, step 18(c~ medical database analysis and matching segments step 18(b) with database collection o similar segments and categorizing the same; comparing step 18(d) said portions to accepted treatments and diagnosis in a database; and grouping step lBte) the iets of treatments and diagnosis with corresponding segments.
once a complete comparlson database step 18 is formed, data from a patient examined may be isolated step l9(a), the data sent to the same database step l9~b);
statistical (a~ compared with medical) database analysis and matching p~ttern recognition step 19(c) oP the portions stzp l9(a) o~ the control chart s~t limits comparing the isolated portion to similar portions o~
control charts in the database, step 19 (d) comparing the isolated portion or se~nent to the database collection of similar se~nents and categorizing the same and finally displaying step l9(e) diagnosis and treatment information from matched portions from the database to the user. ~he display may include the implementation of treatment.
Readjusting control limits step 20 comprises the readjustment of the control chartQ to changed conditions by repeating the steps 1 through 9 above as the patient's condition varies requiring new control charts.
Figures 1 through 9 show how in~ormation is inter-pretable either by a person monitoring the device or by electronically monitoring the device. In each Figure 1 '; " ' ,' ~ W091~10395 2 ~ ~ 0-3 2 ~ PCT/US9ltO025~
through 9, the top Chart, e.g. Figure l(a) shows a normalized patient.
The average avexage, sigma bar and con~rol limits for all three Charts, e.g. Figure l(a), Figure ltb) and Figure l(c), have been set according to these norms~
That is steps 9-11.0 and 12 thro~gh 16 have been done only one time to formulate all three charts in each Figure.
The bottom Charts i~ each Figure, e.g. Figure l(b) and Figure l(c), show the same chart where the normalized patient control chart is still being used but either ta) the range ~or the average has been changed or (~) the range ~or the standard de~iation has been changad slightly or (c) the range ~or the average and standard deviation have been changcd slightly. ~hese charts are arti~icially produced, but the same results are available ~rom patient studies with the invention.
Analysis o~ all charts is similar. For purpases o~
the discussion, only Charts on Figure 1 are specifically discussed. The same analysis applies to Figures 1 through 9.
The c~art features, generated in steps 9 through 15 are given numerically as well as graphically. The chart shows the upper control limit (three sigma is used) 32 which i~ three sigma above ;he average ave~age middle line 33 on ~he chart representing the average average or x-bar from step 14(a~. ~he upp~r control limit numeri-cally displayed 32 is graphically displayed as a UCL line 2~ 3~
WO91/l0395 PCT/U~1/002 4B. The average average middle line 33 iis displayed as the avera~e line 49. The lower control limit numerie display 34 i5 displayed as a LCL line 50 opposite ~he average line 49 from the UCL line 48.
Below the x chart described above is the sigma chart. The sigma chart has the upper control limit 35 (three sigma from step 14(b) above the centerline 36), t~e center 36 is sigma bar from step l~(b), the lower control limit is 37 is opposite the center line 36 from the upper control limit 35. The display shows a UCL line 51 for the upper control limi~ 35, a LCL line for the lower control limit 35 and a center line 52 ~or the center 36.
The blank ~7 shown in Figur~ l~a) runnin~ perpendic-ular to the center line~ o~ both charts ~epre~ents the index location 17(a) o~ the plotter, where the next point is to be plotted. Along the bottom are specification for the number af subgroups 1 plotted, zero 38, ten 39, twenty 40, thirty 41, forty 42, fifty 43, and sixty 44.
The top irregular line 45 represents the graphed appear-ance of the average of subgroups 1. ~he bottom irregular line 46 in Figure l(a) represents the appearance of sigma for subgroups 1.
To the side of each of the charts in Figure 5 are specifications for Average and Standard Deviation.
Specified Average 26 and specified Standard devia~ion 27 are used to generate charts appearing in Figures l(a), l(b) and l(c). The top irregular line 45 and botto~

.:.. .: ~ ' , - . . . ,:

~ WO91/1039~ 2 0 ~ ~ 3 2 ~ PCT/~ 2~
irregular line 46 in Figure l(a) also uses specified averaqe 26 and specified standard deviation 27. In Figure ltb), top Lrre~ular line 45 a~ bottom irregular line 46 use specified average eleven 28 and speci~ied standard deviation 29. In Figure l(c), top irregular line 45 and bottom irregular line 46 use specified average twelve 30 and specified standard deviation two 31.
Charts i~ Figure l(a) represent the normalized patient Figures l(b) and l(c) show the function of the invention as the patient varies from the normalized state. ~igh point 54 shows a subgroup having an average (x-bar) f~om step ll(a) greater than the upper control limits ~et in step l4~a) and there~ore marked by a cirale. Run point 55 sh~ws a subgroup l which ~ollows a number oP subgroups which consecutively are above the a~erage line 52 without a break ~or a point below the average line 52.
Circled points on Figures l through 9 show where the readings have exceeded either the control limits (above the dotted lines) or have exceeded the theory of runs, more than a certain number of readings on one side o~ the average.
Figures l(a) through 9(a) show similar normalized charts. Deviations as indicated on the respective charts are shown in Figures l(b~ through 9(b) and 1(c) throu~h 9(c). These are provided in an ~ort to show the product resulting ~rom the use of the invention.

WO gl/10395 ~ 3 2 ~ PC~/~Sg~
~8 Although the sigma factor is usually the same for the upper and lower control limits on either chart, it may be varied for each separately in step l(a) without departing from the inventive concept and this may prove desirable for certaln situations. The average upper control limit 48, the lower average contr~l limit 50, and the upper sigma control limit 51 are displayed as dotted lines. The lower sigma control limit 53, x-bar line ~5, and sigma bar line 5~ are d~splayed as a solid line. The dotted line is merely a helpful method of display and solid lines may also be dotted. The lines may be diæplayed in dif~erent colors, shapes, etc. without departing ~rom the original concept.
Each o~ Figures 1 - 9 show thrae sets o~ two control chart6. The two con~rol charts in th~ u,ppe~ sot o~ each ~gure shows the normalized pa,tient graph as set up using steps 1 - 17 above. The average and standard deviation used, in establishing these charts is indica,ted at the right o~ the chart for instructional purposes.
The next two charts in each of Figures 1 through 9 show whi~t is displayed when ~sing the same charts but varying average and standard deviation as shown to the right. The step 17 is applied only to the existing control charts. The statistically si~nificant variations are shown as circled m~rks in the display. Although any number o~ va,riations is possible on a given pa~ien~, these displays are instructional as to showing what ma~
occur given various changes.

2~32~
WO9~/10395 PCT/US5~ 2 It i5 an additional improvement over the prior art arising from the disclosure that different confidence levels, upper and lower contrul limits and theory of runs, may be used at once. As shown in Table 2, 6tep 5~
These confidence levels could be used so that when the first control limit was reached below or above the midline, an alarm would sound; and when the second control limit was reached, ~arther from the midline than the ~irst, action would be dictated. Similarly, informa-tion could be sent to the database for comparison after the first was reached, and treatments and diagnosis displayqd upon reaching the second.
one use of the techni~ues developed here~n would be for moni~oring the e~ectiv~ness o~ displays on ~qulp-ment. Either using separate equipment with the e~uipment to be tested or merely using the equipment to be tested, a statistical analysis of the equipment to be tested could be condu~ted. In this way, on a single time line, the stati~tical in~ormation from the control chart could be plotted along with the display information ~rom the equipment to be tested. A comparison of the two would show the effectiveness of the equipment.
Because many varying and different embodiments may be made within the scope of the inventive concept herein taught and because many modifications may be made in the em~odiment(s) herein detailed in accordance with the descriptiva requirements of the law, it is to be un~erW
stood that the details herein are to be interpreted as 2(~32~ ~
W~ 91/t~39 PCr/US91~02 illustrative and not in a limiting sense.

:
` ~ ~

~; .

Claims

IN THE CLAIMS
1. The process pf monitoring patient vital signs from a monitoring device generating data using control charts with at least one control chart limit comprising the steps of:
(a) collectioning said data from said monitoring device.
(b) placing said data into statistically significant subgroups of at least one datum each;
(c) calculating for said statistically significant subgroups statistics to graph against said control charts;
(d) repeating the process steps a though c continu-ously:
(e) selecting for a statistically significant number of repetitions of steps a through c the data necessary to set at least one control chart limit;
(f) setting at least one control chart limit with said data;
(g) setting up at least one control chart with said at least one control chart limit;
(h) continuously graphing said statistics against said at least one control chart.
2. The process of Claim 1 wherein the statistically significant subgroups are consecutive.
3. The process of Claim 1 wherein the control chart limit comprises at least one upper control limit.
4. The process of Claim 1 wherein the control chart limit comprises at least one lower control limit.
5. The process of Claim 1 wherein the control chart comprises at least one midline.
6. The process of Claim 4 wherein the control chart has at least one lower control limit.
7. The process of Claim 1 wherein the control chart has a midline, at least one upper control limit and at least one lower control limit.
8. The process of Claim 1 wherein the process includes the additional step of (i) marking statistically significant deviations.
9. The process of Claim 1 wherein the statistics comprise the average and the standard deviation (sigma) for each of the statistically significant subgroups.
10. The process of claim 10 wherein in the step of selecting for a significant number of repetitions the data necessary to set said control chart limits comprises the additional step of calculating the average of the averages (x-double bar) and the average of the standard deviations for said statistically significant number of repetitions.
11. The invention of Claim 1 comprising the additional step of:
(i) readjusting the control chart limits by repeat-ing the steps a - g for originally setting the control chart limits as set forth above as the patient's condi-tion varies requiring new control charts.
12. The process of Claim 1 comprising the addi-tional steps of:
(i) sending data from a first isolated portion of a first control chart to a database;
(j) categorizing said first isolated portion to the database collection of similar segments and categorizing by similarity;
(k) grouping said first isolated portions with medical data in said database;
(l) comparing said first isolated portion in said database to second portions of a second control chart so as to match said second control chart portion to the first database portion; and (m) displaying the accepted medical data with said portion first.
13. The process of Claim 1 wherein step a further comprises the steps of:
(i) selecting data from the stream of data originat-ing from the monitoring device;
(ii) compartmentalizing data into records;
(iii) isolating that datum of each selected record related to the monitored vital sign.
14. The process of Claim 1 wherein the graphing further comprises matching each subgroup with an associated time.
15. The process of claim 1 wherein the device has a device display and comprising the additional step of comparing the display of the monitoring device against the control charts to determine the reliability of the monitoring device display.
16. The process of Claim 6 wherein there is a first and second different control limit on a single side of the control chart, midline being a first control limit and second control limit and wherein one of the two control chart limits is farther out than the other control limit.
17. The process of Claim 17 comprising the addi-tional step of giving a warning signal is given when the first control limit is reached.
18. The process of Claim 18 comprising the addi-tional step of directing a separate action when the second limit is reached.
The process of Claim 17 wherein the information is sent to a database after the first control limit is reached.
20. The process of Claim 1 wherein step g, setting up at least one control chart, further comprises:
a. at least one control chart from the following set:
(i) Average charts (ii) range charts (iii) P charts, setting out the percent effec-tive or defective (iv) U charts (v) C charts (vi) NP charts - number of occurrences (vii) sigma charts.

21. The invention of Claim 20 wherein the sigma and average chart are used together and wherein the invention comprises the additional steps of:
a. determining at least one of the following standards: the size of the sample subgroup, the method for introducing delay, the number or set of subgroups, the statistical method to be used, the confidence level for entering limits, the confidence level for action limits, setting the theory of runs.
22. The process of Claim 20 wherein steps (a) through (h) further comprise:
i. inputting of data;
ii. compartmentalization of data into records:
iii. grouping of preferably consecutive data into subgroups;
iv. finding an average (x-bar) and finding the standard deviation (sigma) for each subgroup;
v. storing the x-bar and sigma;
vi. Repeating steps i through ii continuously with a delay selected by the user;
vii. selecting a statistically significant number of subgroups;
viii. calculating control chart values for an average (x-bar) control chart for a sigma control chart:

ix. generating two charts, an X-bar chart or average chart, and a sigma chart;
x. continuously graphing information from step s on the charts.
23. The invention of Claim 21 wherein range is substituted for the sigma chart and a range chart is substituted for the sigma chart.
24. The process of Claim 14 further comprising the step of graphing the control chart using a time dependent scale to mark the distance between markings. 2 5 .
The invention of claim 20 wherein the step of determin-ing the average-average, x-double bar, and sigma is replaced with the step of specifying x-double bar and sigma bar.
26. The invention of Claim 25 comprising the additional step of stabilizing the patient within a set range and deviation which the user would select medically specified average and sigma values. 27. The inven-tion of Claim 1 wherein the entire process is used for at least two of the following data types known in the art:
variable data, binomial data, and percentage data from the processes analyzed.
28. The process of Claim 12 wherein the medical data comprises treatment.
29. The process of Claim 12 wherein the medical data comprises diagnosis.
30. The invention of Claim 20 wherein the step of setting up a control chart limit comprises the additional steps of identifying all possible instrumentation and non-medical outliers and eliminating the causes if possible establishing control charts free of the identi-fied outliers and eliminating all possible instrumenta-tion and non-medical outliers.
31. The invention of Claim 1 wherein the chart features generated are shown numerically.
32. The invention of Claim 31 wherein chart and the numeric display comprises;
a. at least one of either a sigma chart or a range chart or an x-bar chart;
b. numeric displays of the upper and lower control limit and midline the same being displayed numerically beside the graphical display for both the x-bar chart and the sigma chart.
33. A method of monitoring a patient's vital signs comprising the steps of:
(a). examining a patient to get data;
(b). isolating data Prom a patient:
(c). sending the data to a database;
(d). statistically (as compared with medical) analyzing the data relative to the database;
(e). matching pattern recognition of the portions of the control chart comparing the isolated portion to similar portions of control charts in the database:
(f). comparing the isolated portion or segment to the database collection of similar segments and catego-rizing the same; and (g). displaying diagnosis and treatment informa-tion from matched portions from the database to the user.
34. The invention of the claim 33 comprising the additional step of implementation of treatment displayed.
35. A method of analyzing data from new medical patients comprising the following steps:
a. Determining number of consecutive runs the theory of runs;
b. Determining factor times sigma for an X-average chart, standard deviation chart, and range chart for upper and lower control limits;
c. Determining number of data samples to be used in each sub-group:
d. Determining number of repetitions for sampling for determining the average standard deviation on midline for the sigma, range, and average control charts;
e. Determining sample rates;
f. Inputting data;
g. Compartmentalizing data into records;
h. Isolating significant portions of each record for graphing on control charts;
j. Selecting consecutive subgroups of statistical-ly significant size;
k. Calculating the average, standard deviation, and range for each subgroup;
l. Pausing as necessary to avoid co-variance between subgroups:
m. Repeating steps of selection, calculating, and pausing for a number of repetitions set by the user in step (d) above;

n . Averaging the averages over all subgroups, standard deviation over all subgroups, and range or standard deviation over all subgroups for the number of repetitions;
o. Setting control limits by multiplying the B
factor times the average sigma in order to determine the upper and lower control limits;
p. Set up control charts utilizing the averages for averages, standard deviations, and ranges calculated above;
q. Continuously repeating the steps of input of data, compartmentalization of data, isolation of signifi-cant portions of records, selecting consecutive sub-groups, calculating the average, standard deviation, and range;
r. Continuously graphing on or against a control chart and displaying on a control chart each of the consecutive subgroups:
s. Marking deviations as to the theory of runs as set out above, subgroups whose statistical information goes over the control limits;
t. Readjusting the control limits as desired;
u. Isolating segments of the control charts;
v. Comparing isolated portions of the control charts to similar isolated portions of the control charts in a database;
w. Displaying or applying treatments for diagnosis attached to the matched similar charts in the database.

36. The process of Claim 1 wherein the statistics comprise the average and the range for each of the statistically significant consecutive subgroups.
37. The process of claim 36 wherein in the step of selecting for a significant number of repetitions the data necessary to set said control chart limits comprises the additional step of calculating the average of the averages (x-double bar) and the average of the ranges for said statistically significant number of repetitions.
38. The process of Claim 36 further comprising the step of displaying the results on the time dependent scale so that each subgroup is matched with the associat-ed time.