the development of synoptic reports from free text content of archival pathology reports generated...

17
THE DEVELOPMENT OF SYNOPTIC REPORTS THE DEVELOPMENT OF SYNOPTIC REPORTS FROM FREE TEXT CONTENT OF ARCHIVAL FROM FREE TEXT CONTENT OF ARCHIVAL PATHOLOGY REPORTS GENERATED IN THE PATHOLOGY REPORTS GENERATED IN THE ANATOMIC PATHOLOGY LABORATORY ANATOMIC PATHOLOGY LABORATORY INFORMATION SYSTEM INFORMATION SYSTEM

Upload: brent-malone

Post on 18-Dec-2015

217 views

Category:

Documents


0 download

TRANSCRIPT

THE DEVELOPMENT OF SYNOPTIC THE DEVELOPMENT OF SYNOPTIC REPORTS FROM FREE TEXT CONTENT REPORTS FROM FREE TEXT CONTENT OF ARCHIVAL PATHOLOGY REPORTS OF ARCHIVAL PATHOLOGY REPORTS

GENERATED IN THE GENERATED IN THE ANATOMIC PATHOLOGY LABORATORY ANATOMIC PATHOLOGY LABORATORY

INFORMATION SYSTEMINFORMATION SYSTEM

Purpose of Synoptic Purpose of Synoptic ReportsReports

Structure the informationStructure the information Structured form facilitates Structured form facilitates

research queriesresearch queries Structured form facilitates better Structured form facilitates better

patient carepatient care Structured form facilitates better Structured form facilitates better

sharing of information between sharing of information between institutionsinstitutions

Example Synoptic Example Synoptic PagePage

Kidney – Genitorurinary CAP Approved Surgical Pathology Cancer Case Summary (Checklist)

Protocol revision date: January 2005 Applies to invasive carcinomas only

Based on AJCC/UICC TNM, 6th edition

KIDNEY: Nephrectomy, Partial or Radical Patient name: Surgical pathology number: Note: Check 1 response nuless otherwise indicated. MACROSCOPIC Speciman Type ___Partial nephrectomy ___Radical nephrectomy ___Other (specifiy): ________________________________ ___Not specified Laterality ___Right ___Left ___Not specified *Tumor Site (check all that apply) *___Upper pole *___Middle *___Lower pole *___Not specified Focality ___Unifocal ___Multifocal Tumor Size (largest tumor if multipled) Greatest dimension:___cm *Additional dimensions:___x___cm ___Cannot be determined (see comment)

Conversion of Archived Conversion of Archived SystemsSystems

An institution that changes from free An institution that changes from free formatted pathology reports to formatted pathology reports to synoptic reporting may wish to synoptic reporting may wish to convert all archived reports to convert all archived reports to synoptic formsynoptic form

Is it better to convert archived data by Is it better to convert archived data by hand or automatically with respect to:hand or automatically with respect to:• AccuracyAccuracy• TimeTime

Computer MethodsComputer Methods

Search for key wordsSearch for key words Use auto-coding vocabularies so Use auto-coding vocabularies so

that all (or most) variations of that all (or most) variations of linguistically equivalent phrases linguistically equivalent phrases are placed in a uniform phrase for are placed in a uniform phrase for proper pattern matching in proper pattern matching in searchessearches

Use negation algorithmsUse negation algorithms

Special ChallengesSpecial Challenges(What is the Tumor Site for this case?)(What is the Tumor Site for this case?)

Part A. RECTOSIGMOID COLON ULCER, BIOPSY:Part A. RECTOSIGMOID COLON ULCER, BIOPSY:--INFILTRATING MODERATELY DIFFERENTIATED --INFILTRATING MODERATELY DIFFERENTIATED

ADENOCARCINOMA, CONSISTENT WITH ADENOCARCINOMA, CONSISTENT WITH COLONICCOLONIC

ORIGIN, SEE NOTE.ORIGIN, SEE NOTE.

Part C. COLON, LEFT, HEMICOLECTOMY:Part C. COLON, LEFT, HEMICOLECTOMY:--INFILTRATING MODERATELY DIFFERENTIATED --INFILTRATING MODERATELY DIFFERENTIATED

ADENOCARCINOMA OF RECTOSIGMOID ADENOCARCINOMA OF RECTOSIGMOID COLON, COLON,

SEE NOTE.SEE NOTE.

Which is “Whipple” and which Which is “Whipple” and which is “Whipple Pylorous Sparing”?is “Whipple Pylorous Sparing”?

Received fresh for intraoperative consultation, labeled Received fresh for intraoperative consultation, labeled with the patient's name, number and "with the patient's name, number and "WhippleWhipple," is a ," is a bloc section of organs from a Whipple procedure. The bloc section of organs from a Whipple procedure. The specimen consists of a 24.0 cm. in length duodenum specimen consists of a 24.0 cm. in length duodenum and a 6.2 x 4.3 x 4.0 cm. remnant of attached and a 6.2 x 4.3 x 4.0 cm. remnant of attached pancreas, and a portion of distal pancreas, and a portion of distal stomachstomach, 6.5 x 2.5 , 6.5 x 2.5 cm.cm.

Received fresh for intraoperative consultation, post-Received fresh for intraoperative consultation, post-fixed in formalin, labeled with the patient's name and fixed in formalin, labeled with the patient's name and number, is an en bloc resection of organs from a number, is an en bloc resection of organs from a WhippleWhipple procedure. The specimen consists of a 21.5 procedure. The specimen consists of a 21.5 cm. length of duodenum, a 1.5 cm. length of distal cm. length of duodenum, a 1.5 cm. length of distal common bile duct, and an 8.8 x 6.8 x 6.7 cm. remnant common bile duct, and an 8.8 x 6.8 x 6.7 cm. remnant of attached pancreas. of attached pancreas.

Design of ExperimentDesign of Experiment

To determine the feasibility of automated To determine the feasibility of automated conversion of archived reports, we collected all conversion of archived reports, we collected all records for the year 2005 and converted them records for the year 2005 and converted them to synoptic form, both manually and by to synoptic form, both manually and by computer text processing. We then compared computer text processing. We then compared the results for accuracy and the time to process. the results for accuracy and the time to process. The organ sites we studied were: The organ sites we studied were:•PancreasPancreas•KidneyKidney•ColonColon•ProstateProstate•BreastBreast

Estimated Accuracy for Estimated Accuracy for PancreasPancreas

SpecimeSpecimen typen type

Tumor Tumor SiteSite

Tumor Tumor sizesize

HistologiHistologic typec type

HistologiHistologic Gradec Grade

23/2523/25

92%92%

18/2518/25

72%72%

19/2519/25

76%76%

24/2524/25

96%96%

24/2524/25

96%96%

Estimated Accuracy for Estimated Accuracy for KidneyKidney

LateralityLaterality Tumor Tumor SiteSite

Tumor Tumor SizeSize

HistologiHistologic Typec Type

27/2727/27

100%100%

21/2721/27

78%78%

7/277/27

26%26%

24/2724/27

89%89%

Estimated Accuracy for Estimated Accuracy for ColonColon

Tumor SiteTumor Site Tumor SizeTumor Size Histologic Histologic TypeType

40/7540/75

53%53%

42/7542/75

56%56%

74/7574/75

99%99%

Estimated Accuracy for Estimated Accuracy for ProstateProstate

HistologiHistologic Typec Type

Gleason Gleason PrimaryPrimary

Gleason Gleason SecondarSecondaryy

Gleason Gleason TotalTotal

49/5049/50

98%98%

48/5048/50

96%96%

48/5048/50

96%96%

48/5048/50

96%96%

Estimated Accuracy for Estimated Accuracy for BreastBreast

Bloom-Bloom-Richardson Richardson gradegrade

Mitotic ScoreMitotic Score AngiolymphatiAngiolymphatic Invasionc Invasion

100%100% 100%100% 100%100%

Estimated Time for Estimated Time for Computer ConversionComputer Conversion

The average time to develop The average time to develop the text processing code was the text processing code was approximately 4-6 hours per approximately 4-6 hours per organ. The first organ took organ. The first organ took the longest time, the others the longest time, the others were re-written from the first.were re-written from the first.

Estimated Time for a Trained Estimated Time for a Trained Pathologist to convert archive Pathologist to convert archive

reports to synoptic reportsreports to synoptic reports

Total Total TimeTime

Number Number of Casesof Cases

Average Average TimeTime

PancreasPancreas N/AN/A N/AN/A N/AN/A

ColonColon 9:27:489:27:48 7272 7:537:53

BreastBreast 4:46:464:46:46 4848 5:585:58

KidneyKidney 3:57:233:57:23 3232 7:257:25

ProstateProstate 3:38:303:38:30 3030 7:177:17

ConclusionConclusion

Less than 40-60 cases, a trained Less than 40-60 cases, a trained pathologist is quicker.pathologist is quicker.

More than 40-60 cases, automated More than 40-60 cases, automated text processing is quicker.text processing is quicker.

Attempts to improve the accuracy will Attempts to improve the accuracy will increase this time.increase this time.

For archived systems with a very long For archived systems with a very long history and a large volume of patients, history and a large volume of patients, automated processing may be automated processing may be preferred.preferred.

AcknowledgementsAcknowledgements

We wish to thank the We wish to thank the following for their following for their participation in this project:participation in this project:•Ashok PatelAshok Patel•Doug HartmanDoug Hartman•John GilbertsonJohn Gilbertson•Robert LaneseRobert Lanese•Rajnish GuptaRajnish Gupta