Last Updated: 02/25/2011

NEXT EMAIL
From: Murphy, Shawn N. Sent: Friday, February 25, 2011 8:28 PM To: Jack London Cc: i2b2 AUG Members
Subject: RE: Release schedule
Hi Jack,
No, the UI for using the modifiers will only be in the May release.
Thanks, Shawn.

From: Jack London Sent: Friday, February 25, 2011 5:20 PM To: Murphy, Shawn N. Cc: i2b2 AUG Members members
Subject: Re: Release schedule
Will RC3 fully utilize the concept_cd modifiers?

On Feb 25, 2011, at 5:16 PM, Murphy, Shawn N. wrote:
Hi Keith,
We are targeting the RC3 release of 1.6 for next week. There is still a lot to be done before the final release of 1.6, so May 2011 is our current target.
There will be a 1.7 release, likely in the Spring of next year. This release will focus on the connections with Enterprise identity management, import services that can hook up to the EMR and other external sources, and further query enhancements.
Ontology and Web Client work will accompany both releases.
We are backing off from frequent production releases, in favor of more frequent Release Candidates. The reason is that the burden of updating in production is high, and so it is better to target them for once a year or so. However, the most current work is available in the Release Candidates if a team wishes to implement some features early.
Thanks,
Shawn.

From: Keith Marsolo Sent: Friday, February 25, 2011 3:51 PM To: i2b2 AUG Members
Subject: Release schedule
Shawn and/or Mike,
Three questions:
1. Do you guys have any kind of ETA on 1.6RC3?
2. What's the best guess on the release date for the final candidate of 1.6?
3. Is there still a plan for version 1.7 or will 1.6 be the last core version?
Not trying to apply any pressure, just curious.
Thanks.
Keith


NEXT EMAIL
From: Sebastian Mate Sent: Thursday, February 24, 2011 11:27 AM To: i2b2 AUG Members
Subject: Let the user change his/her i2b2 password
Hi everybody,
I know this came up some months ago, but as far as I remember, there was no solution.
Is there a way to enable the normal i2b2 user (without the "ADMIN" role) to change his/her password? I know the old gridsphere portal in 1.3 had this option, but it seems to be missing in the recent web client / admin. This is something that we are really missing. Whenever we create a new user, the password defaults to "demouser". The first question is always: "How can I change this"?
If there is no solution so far, I will try to find a solution. Maybe a small webservice which calculates the MD5 hash will do the trick, but I don't want to reinvent the wheel, maybe I am missing something ...
Thanks a lot in advance, Sebastian
Dipl.-Inf. Sebastian Mate Chair of Medical Informatics, Friedrich-Alexander-University Erlangen-Nuremberg, Germany Krankenhausstrasse 12, 91054 Erlangen


NEXT EMAIL
From: Mark Weiner Sent: Wednesday, February 23, 2011 12:56 PM To: 'Arvinder Choudhary'; 'Dan Connolly' Cc: i2b2 AUG Members; alicel
Subject: RE: I2B2 speed of lab queries and thin client plug ins
Thanks for these suggestions.
Recreating the table with fewer numbers of larger extents turned out not to be helpful.
Our Block size is 8K, and this may have been another OS limitation out of our control when the cloud machine was configured.
However, one thing we just did was to analyze the indexes (fact_cnpt_pat_enct_idx for starters) as well as the observation_fact table. That seemed to enable these lab value queries to run in < 60 seconds whereas before they were timing out. I also got brave and set the PGA_AGGREGATE_TARGET value to 2G which also seemed to help – making the queries run in < 30 seconds.
Its hard to tell what is actually working because with caching, rerunning queries is always faster. However, before, when I altered the query by using a different lab test, the query seemed to take as long as the first time. Now the query is faster even when I alter the lab test that I am querying.
"Not timing out" on basic queries is an important end user requirement!
Mark

From: Arvinder Choudhary Sent: Wednesday, February 23, 2011 10:57 AM To: Dan Connolly; Mark Weiner Cc: i2b2 AUG Members; alice l
Subject: RE: I2B2 speed of lab queries and thin client plug ins
Mark:
If you think the small extent is the issue.. There is a limit for each database datafile's maximum size, without specifying BIGFILE keyword, the limit is 222 Blocks (which is 4 Million blocks), if the block size is 8k, the file limit will be 32768MB.
With BIGFILE keyword, the limitation is 4 billion blocks..then try this ...when you create the tablespace and your block size is 8K than you can create tablespace bigger than 32GB you mention BIGFILE in the create tablespace script..
if your blocksize is 16 or 32 ...than you can don't have to mention BIGFILE in tablespace creation to have over 32GB
say for e.g
create BIGFILE tablespace abc ....... BLOCKSIZE 8K ...
instead of just
create tablespace abc ....
I hope this helps
Regards
Arvinder Choudhary
BE MS PMP Medical informatics Project Director University of Kansas Medical Center 3901 Rainbow Blvd, KS 66160

"Mark Weiner" 2/22/2011 4:34 PM
Thanks Dan (and James and Deb).. I dont think the hardware is the limiting issue. It is a cloud machine simulating a dual quad core with 32 GB RAM and 1TB disks, though I dont have a sense for how striped the data is on the disks. Interestingly, when the machine was configured, it was set up with a 32 bit OS, so our data files are limited to 32 GB. Hopefully that little snafu is not the key factor in the query speed.
We have reindexed and re-analyzed without much effect on speed. One thing we are looking at is the fact that our observation fact table was expanded from the original table definitions that had relatively small extent sizes. So our 100+GB table was split into many many small extents, which may be slower than a few bigger extents.
Can anyone compare notes on potentially relevant Oracle tuning parameters?
Mark

From: Dan Connolly Sent: Tuesday, February 22, 2011 12:12 PM To: m weiner Cc: i2b2 AUG Members; alice l
Subject: Re: I2B2 speed of lab queries and thin client plug ins
I can't see an obvious problem or fix, but I can share some relevant experience...
We have about half that many lab facts. I think our total row count is in the same ballpark as yours.
From a pretty cold state, a query by lab value for sodium took 40 seconds here. Hmm...it was a 0-patients query, so perhaps that's not helpful... OK, I ran a different query one and got ~60K patients. i2b2 reports 44.6 seconds.
Perhaps you're exhausting RAM? That doesn't seem terribly likely. We have 8GB in our database server, which is pretty ordinary. And we use reasonably fast disks and RAID 1+0.
Maybe an index into an unusable state or something? That happens to us when we load data sometimes, though it results in errors at query time, not poor performance.
We have some hopes of doing plug-in development, but we haven't done any yet. We have done a little debugging of the existing plug-ins.

On Mon, 2011-02-21 at 13:48 -0500, Mark Weiner wrote:
Happy President's day!
Our I2B2 instance has about 150 million rows of which 52 million are labs. Concept-based queries are relatively fast taking no more than a few seconds,or up to 20-30 seconds for those that involve intersections of different criteria. However, queries by lab value typically exceed the 3 minute timeout, and while I could increase the interval to timeout, I'm more interested in ways to make the query run faster. Typically after a timeout,a rerun of the same query goes faster, but that is not the way we want our instance to run.
I dont have a sense of how pervasive the slow lab value query is among theI2B2 user group. Have others had success with indexing aside from the strategy used in the demo data?
Also, has anyone outside of Partners has success with developing freshplug-ins (or even modifying existing plug ins) or the thin client?
Thanks!
Mark
Mark Weiner, M.D.Director of Information Systems Integration for ResearchOffice of Human Research University of Pennsylvania School of Medicine423 Guardian Drive – Rm 1116 Philadelphia, PA


NEXT EMAIL
From: Murphy, Shawn N. Sent: Thursday, February 17, 2011 8:33 AM To: Peter Beninato; i2b2 AUG Members Subject: RE: i2b2 - observation fact - unit_cd - metadataxml UOM - question
Hi Peter,
Please see a full explanation in the wiki under:
https://community.i2b2.org/wiki/display/DevForum/Metadata+XML+for+Medication+Modifiers
Although it specifically addresses units for medications, it is universally applicable.
Thanks, Shawn.

From: Peter Beninato Sent: Wednesday, February 16, 2011 1:23 PM To: i2b2 AUG Members
Subject: i2b2 - observation fact - unit_cd - metadataxml UOM - question
Hi,
In the c_metadataxml field in the i2b2 metadata table, there are tags for UOM (unit of measure).
There is also a units_cd field in observation_fact.
Are they tied together in any way?
Why?
So looking at labs, a particular lab might have different UOMs for instance: Ug/dl And Mcg/dl
Might like to present the UOM in the drop down list that the c_metadataxml field presents, and then use that UOM to look for observations for that LAB with that units_cd.
Or is there a different approach ( like v1.6 modifier_cd)?
Thanks.
Peter Beninato - OCTRI DW Developer Oregon Health & Science University


NEXT EMAIL
From: Phillip Reeder Sent: Friday, February 11, 2011 5:36 PM To: Keith Marsolo; Dan Connolly Cc: j cowall; i2b2 AUG Members
Subject: RE: Old patient record question...
Agreed. With our outpatient dataset, we have roughly 2 million patients but only ~300K have data. I've limited it to only patients that have data in the EMR.
Phillip

From: Keith Marsolo Sent: Friday, February 11, 2011 4:09 PM To: Dan Connolly Cc: j cowall; i2b2 AUG Members
Subject: Re: Old patient record question...
We have similar patients (demographics, no visit data) in our source systems, but elected to not load them into the warehouse because it can give people inflated expectations about the population size. If people expect that 20% of a population is diabetic and you tell them you have 3 million patients, they're going to assume you've got 600,000 diabetics in your warehouse. If 2 million of those patients are "phantom," when you tell them the actual number is 200,000, they aren't going to be happy (especially if they put it into a grant). Along the same lines, it keeps you out of trouble when certain people (usually above your pay grade) get into "mine's bigger than yours" arguments about warehouse size. You want to make sure you keep the emperor clothed.
That's my two cents anyway.
Keith

On Feb 11, 2011, at 4:27 PM, Dan Connolly wrote:
We're in a similar situation.
We haven't found much motivation to do the extra work of filtering out the older demographic data, especially since we hope to bring in data from more sources over time, and we want to correlate patients between all sources.

On Fri, 2011-02-11 at 16:12 -0500, Jeff Cowall wrote:
Greetings, all.
We have a bit more than 3,000,000 distinct patient identifiers in our source warehouse, but of these, only roughly a million have an encounter of any kind since we start keeping warehouse data some 10-12 years ago. Best we can tell, these identifiers come from the initial backloading of new system at that time. >> >> I'm loading all of it, including demographic observation_fact rows, but I'd like to know the thoughts of the group about the utility of loading old demographic data with no other observations at all.
I'm asking our internal research consultant, too, but thought it useful to ask the group.
Thanks,
Jeff


NEXT EMAIL
From: Murphy, Shawn N. Sent: Thursday, February 10, 2011 1:56 PM To: Dustin Schultz; i2b2 AUG Members Cc: Mendis, Michael E.
Subject: RE: Security Issues
Hi Dustin,
If you could send them to Mike and I that would be great.
Thanks, Shawn.

From: Dustin Schultz Sent: Thursday, February 10, 2011 1:20 PM To: i2b2 AUG Members
Subject: Security Issues
Hi,
I've identified a number of security issues within the i2b2 server code and the interest of responsible disclosure I'd like to know where I can disclose these? Is there a security list I can submit them too?
Best,
N. Dustin Schultz Software Engineer University of Utah Office of the AVP for Health Sci. IT FURTHeR Project


NEXT EMAIL
From: Michael C Ford Sent: Monday, February 07, 2011 6:07 PM To: i2b2 AUG Members
Subject: database tables and the data_OBFSC
Setting up a new environment, Should be working but I find when I run a query I get the following error. something about the setup seems incorrect.
2011-02-07 15:02:21,147 INFO [edu.harvard.i2b2.crc.dao.DataSourceLookupHelper] Located DataSource for hiveId=[i2b2demo] projectId=[/Demo/] 2011-02-07 15:02:21,148 INFO [STDOUT] Roles from get i2b2demo/Demo/i2b2 2011-02-07 15:02:21,148 INFO [STDOUT] Roles from size 4 2011-02-07 15:02:21,148 INFO [edu.harvard.i2b2.crc.dao.OracleDAOFactory] Using datasource java:QueryToolDemoDS 2011-02-07 15:02:21,152 DEBUG [org.springframework.jdbc.core.JdbcTemplate] Executing SQL query [select * from i2b2demodata.qt_privilege where protection_label_cd = ? and plugin_id is NULL] 2011-02-07 15:02:21,152 DEBUG [org.springframework.jdbc.datasource.DataSourceUtils] Fetching JDBC Connection from DataSource 2011-02-07 15:02:21,152 DEBUG [org.springframework.jdbc.core.StatementCreatorUtils] Setting SQL statement
parameter value: column index 1, parameter value [SETFINDER_QRY_WITH_DATAOBFSC], value class [java.lang.String], SQL type unknown 2011-02-07 15:02:21,154 DEBUG [org.springframework.jdbc.datasource.DataSourceUtils] Returning JDBC Connection to DataSource 2011-02-07 15:02:21,155 DEBUG [org.springframework.beans.factory.support.DefaultListableBeanFactory] Returning cached instance of singleton bean 'appType' 2011-02-07 15:02:21,159 DEBUG [edu.harvard.i2b2.common.util.jaxb.JAXBUtil] object.toString() edu.harvard.i2b2.crc.datavo.i2b2message.RequestMessageType 2011-02-07 15:02:21,160 DEBUG [edu.harvard.i2b2.crc.axis2.QueryService] Response in service<?xml version="1.0" encoding="UTF-8" standalone="yes"?> <ns5:response xmlns:ns9="http://www.i2b2.org/xsd/cell/ont/1.1/" xmlns:ns4="http://www.i2b2.org/xsd/cell/crc/psm/1.1/"
xmlns:ns7="http://www.i2b2.org/xsd/cell/crc/psm/analysisdefinition/1.1/" xmlns:ns3 ="http://www.i2b2.org/xsd/cell/crc/pdo/1.1/" xmlns:ns5="http://www.i2b2.org/xsd/hive/msg/1.1/" xmlns:ns2 ="http://www.i2b2.org/xsd/hive/pdo/1.1/" xmlns:ns6="http://www.i2b2.org/xsd/cell/crc/psm/querydefinition/1.1/" xmlns:ns10
="http://www.i2b2.org/xsd/hive/msg/result/1.1/" xmlns:ns8="http://www.i2b2.org/xsd/cell/pm/1.1/">
<message_header> <i2b2_version_compatible>1.1</i2b2_version_compatible> <hl7_version_compatible>2.4</hl7_version_compatible> <sending_application>
<application_name>edu.harvard.i2b2.crc</application_name>
<application_version>1.4</application_version> </sending_application> <sending_facility>
<facility_name>i2b2 Hive</facility_name> </sending_facility> <receiving_application>
<application_name>i2b2_QueryTool</application_name>
<application_version>0.2</application_version> </receiving_application> <receiving_facility>
<facility_name>i2b2 Hive</facility_name>
</receiving_facility> <message_control_id>
<instance_num>1</instance_num> </message_control_id> <project_id>Demo</project_id>
</message_header>
<response_header> <info>Log information</info> <result_status>
<status type="DONE">DONE</status> <polling_url interval_ms="100"/>
</result_status> </response_header> <message_body>
<ns4:response xsi:type="ns4:master_instance_result_responseType" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"> <status> <condition type="ERROR">This operation required privilege : DATA_OBFSC</condition> </status> </ns4:response> </message_body> </ns5:response>
2011-02-07 15:02:39,460 DEBUG [org.jboss.resource.connectionmanager.IdleRemover] run: IdleRemover notifying pools, interval: 30000 2011-02-07 15:02:49,840 DEBUG [org.jboss.ejb.txtimer.TimerImpl] run: [id=
1297108587828,target=[target=jboss.j2ee:service=EJB3,ear=QP1.ear,jar=QP-AnEJB.jar,name=CronEjb],remaining=-1851,periode=60000,active] 2011-02-07 15:02:49,840 DEBUG [org.jboss.ejb.txtimer.TimerImpl] setTimerState: in_timeout 2011-02-07 15:02:49,841 DEBUG [org.springframework.jdbc.core.JdbcTemplate] Executing SQL query [ select * from i2b2hive.CRC_ANALYSIS_JOB where queue_name = ? and status_type_id = ? order by create_date] 2011-02-07 15:02:49,841 DEBUG [org.springframework.jdbc.datasource.DataSourceUtils] Fetching JDBC Connection from DataSource


NEXT EMAIL
From: Russ Waitman Sent: Monday, February 07, 2011 11:02 AM To: Keith Marsolo Cc: i2b2 AUG Members; Karen Blackwell
Subject: Re: i2b2 - WEBCLIENT - Patient Count - Query History - "Less Than 10"
Keith,
Thank you for the reivew of your process. I'll discuss further with our IRB. From my perspective the timeline (if shifted and removing identifiers) would pass muster as non-human subjects research but may require additional privileges from an institutional perspective.
Russ

Keith Marsolo 2/4/2011 8:05 AM
Russ,
We're taking a similar approach here at CCHMC. Like yours, our IRB was okay with giving everyone obfuscated access to the web client. We do the same type of de-identification so they ruled it to be non-human subjects research.
We don't have the thick client deployed here, so the timeline view doesn't become an issue. But I've always had the same concerns and didn't think the timeline would pass muster with our IRB. Plus, with the sparsity/heterogeneousness of our data (mostly administrative data from 2003-2007 and then more data as we implemented Epic), I worry about people making the wrong conclusions. I do believe the timeline view has a place, but I don't think it's necessarily at the "enterprise-level."
With our web client, we're implementing a "chart review" mode which would give users access to a de-identified cohort. Once we load all the data, it would essentially function as a de-identified medical record. So in this use case, we'd share many of the same issues/concerns. The idea is that once someone runs an obfuscated query, they can request "elevated" access to the cohort. They would need to fill out some sort of data use agreement, only have access for a certain time and we'd audit requests, etc. Once version 1.6 is finished (or RC3 is released), we plan to leverage the "patient set as query criteria" functionality to make this happen (essentially forcing the cohort to be part of the query). But this does require a higher-level of access. My goal is really to make it so we don't need to provide MRNs and other PHI to people to do chart review. If we can get enough of the historical data into the de-identified warehouse (including notes), I think this is feasible. I need to include this in version 2.0 of our IRB protocol which I've been meaning to get to for the past four months. If and when I finish, I'm happy to share.
As for pulling out datasets, I demonstrated our extract functionality to our Legal department, and the response was more or less, "Yeah, I don't see that happening." But I think we can manage it eventually.
Keith

On Feb 3, 2011, at 6:28 PM, Russ Waitman wrote:
Hi, I'd just expand on Dan's point as we spent a bit of time today thinking about this and it was an interesting coincidence the same issue was brought up on the AUG today.
There's an intersection of what capabilities the tool provides and interpretation of what it means to have a de-identified dataset versus a limited dataset versus best practices that we want to understand and make sure we can clearly explain to the organizations entrusting us with their patient data and our regulatory oversight. I am curious if there is a consistent perspective on de-identification versus best practices by human subject committees with respect to HIPAA and IRB (ccing Karen from our HSC and John Keighley a biostatistics colleague).
We're taking the approach of using the web client and providing a "lower" barrier for research by letting people look at the web client after we've verified they are qualified faculty (via LDAP flag), and they've completed their human subjects training (via training check), and they have signed an electronic system access agreement. By the time it makes it to the database accessed by i2b2, all patients, are deidentified/scrubbed of identifiers and the dates have all been shifted back in time by up to 365 days. This is both preparatory to human subjects research and since the data is de-identified it can be considered non-human subjects research. The auditing and oversight at this level is as much for the contributing organizations as local HSC/IRB.
Once the user wants to pull data off our system, they will complete a data use agreement which will more specifically describe their project and data fields which will be reviewed and approved but since it's based on de-identified data, is won't need to go through IRB, just our data request oversight process.
Ideally, since the i2b2 user can't take anything off the screen (outlined in our agreement; don't screen print or use a browser cache harvesting program), we'd like to give them as much functionality as possible. The ability to take a set, apply simple statistics to understand the population, and then see the timeline so you can understand the relationships between the data is very informative (aha... these patients got the drug after the procedure X while these have an inconsistent pattern). But, by allowing this am I basically defeating any obfuscation best practice?
It would seem there's an informatics community "best practice" of not allowing the isolation of a cohort less than a certain size (10, 5, 20).
Here's an example:
1. I use i2b2 to define a small cohort of patients who have ever had an inpatient diagnosis with cleft palate. It returns 21 patients as a count.
This count would seem to be very safe from a de-identification perspective and their is no knowledge provided yet to subdivide the patients. The user just has a count.
2. I would then like to understand this population so I run the query with a returned patient set.
I put the set in the simple counts view I see that there are 5 patients with an age between 30 and 40. – so, would that be interpreted as seeing a set with less than 10 individuals in violation of my best practice regarding obfuscation?
I then put the set in the timeline tool to see when they get their diagnosis assigned relative to any other observation. – The concepts in the timeline are not necessarily the same as those used to define the cohort so I could quickly identify small subpopulations of the cohort – Additionally, by looking at a timeline, each patient timeline is a unique signal in appearance.
I could say from a HIPAA perspective are we fine because all identifiers are removed (we've shifted the dates randomly by up to year, thus not providing any dates) so it's still deidentified nonhuman subjects research even though the user can detect smaller subpopulations. But from a best practice perspective, I let the user see very individualized patient signals.
The other possibility is that when one develops and deploys analysis tools, there's a setting which will analyze any dimensions used to stratify the cohort and if set to obfuscate, it won't let th users see any subdivisions <10. For cleft palatte, I couldn't really drill down but on a larger diagnosis like diabetes, I could subdivide the population to a greater degree.
Practically, what are other places letting their users do?
Do you have to have an IRB protocol to use the timeline?
Get another level of approval to be granted higher than obfuscation access?
Do you have to have an data use agreement to use the thick client and pull out datasets?
Do you not care and let the approved users fish away since it's a fairly small audience?

Russ Waitman Associate Professor Director of Medical Informatics Department of Biostatistics University of Kansas Medical Center

Peter Beninato 2/3/2011 4:16 PM
For clarity, this is in reference to the Query Status section of the webclient.

From: Peter Beninato Sent: Thursday, February 03, 2011 2:15 PM To: 'Dan Connolly' Cc: i2b2 AUG Members
Subject: RE: i2b2 - WEBCLIENT - Patient Count - Query History - "Less Than 10"
Part of the issue is that when you click "Run Query", a user who has DATA_OBFSC role has the ability to check "Patient Set (list of matching patients)."
When the query is executed, there is a line for "Patient Set" as well as "Patient Count."
I think this is related to the use case question Rob Schuff is trying to ask in the email thread sent today, 2/3/2011 with the Subject: user roles documentation?.
Should someone with the DATA_OBFSC role be able to select "Patient Set (list of matching patients)" from the "Run Query" command?
When they do, without changing the Javascript, the Patient Set count list two values (in addition to the Patient Count line). One of counts for the variable "t" does not seem to be obfuscated.
Fyi.

From: Dan Connolly Sent: Thursday, February 03, 2011 1:41 PM To: Peter Beninato Cc: i2b2 AUG Members
Subject: RE: i2b2 - WEBCLIENT - Patient Count - Query History - "Less Than 10"
We're dealing with pretty much the same issue, and I discovered that i2b2 1.4 will do the "less than 10" thing, i.e. it will say "Obfuscated - 0 patients" for result sets smaller than 10, but only if the user's permissions are DATA_OBFSC only. This is documented in the CRC Design Document:
Data Protection Track Role Access Description DATA_OBFSC Access to aggregate result like the patient counts after it is obfuscated. DATA_AGG Can see the aggregate result like the patient count without obfuscation.
We're considering modifying i2b2 to enforce the "less than 10" stuff even for DATA_AGG and DATA_OBFSC permissions because we'd like to give users access to the timeline and such... but then we realized: you could put a set of 15 patients in the timeline, add another concept where only some of the patient have that sort of observation fact, and effectively display a patient set smaller than 10.
So we're not so sure that's a good idea after all.
If any of the i2b2 designers/developers can share any design history/rationale for this i2b2 feature, I'd appreciate it. Perhaps you've explored many or all of the options that I'm working through and have found requirements that rule out all the designs except the way i2b2 currently works.
Dan Connolly, Biomedical Informatics Software Engineer University of Kansas Medical Center
Division of Informatics

On Tue, 2011-02-01 at 16:49 -0800, Peter Beninato wrote:
Wow.... Left to your own devices...
So it seems that CRC_SDX_QI.js holds promise. Since both USER and DATA_OBFSC are assigned it seems that a similar edit to below:
if (i2b2.PM.model.userRoles.length == 2 && i2b2.PM.model.userRoles[0] == "DATA_OBFSC")
resolves issue for user with DATA_OBFSC role.
I think for DE-Identified users, then the ELSE path for title= will need to be edited to something like "Less Than 10 Patients".
FYI.
Peter

From: Peter Beninato Sent: Tuesday, February 01, 2011 4:00 PM To: i2b2 AUG Members
Subject: i2b2 - WEBCLIENT - Patient Count - Query History - "Less Than 10"
Hi,
Trying to work through some issues with Patient Count in the Previous Queries section (CRC History) in version 1.4 with the WEBCLIENT
When a query is run, there is some unwanted behavior in the Query Status and the Previous Queries panes when the Patient Record Count is LESS THAN 10.
When a query is run if the record count is less than 10, the desired behavior is to display "Less Than 10 Patients."
The issues in the Query Status can be resolved by changing code in CRC_ctrlr_QryStatus.js
i2b2.PM.model.userRoles.length == 2 (for USER and DATA_OBFSC and then modify the lines
title = t+... as needed in the PATIENTSET and PATIENT_COUNT_XML sections
How can the Patient Count be displayed as "Less Than 10 Patients" in the Previous Queries section?
When a Previous Query is expanded via the control and record.size is < 10, it seems an obfuscated Patient Count from the response XML is displayed. <set_size>5</set_size> Is displayed as "Patient Count – 5 Patients"
Can it be changed to "Patient Count - Less than 10 Patients"?
Where?
Could really use some help resolving this issues. Thanks.
Peter Beninato - OCTRI DW Developer Oregon Health & Science University


NEXT EMAIL
From: Murphy, Shawn N. Sent: Thursday, February 03, 2011 5:18 PM To: Robert Schuff; i2b2 AUG Members
Subject: RE: Obfuscated user role....WAS: Re: user roles documentation?
Hi Rob,
Two use cases for seeing the patient set in "previous queries" while the user is in the obfuscated role is:
1 - it may be submitted in a request to get an IRB approved data mart created (in which the the user may be able to see data at a higher privacy role than "obfuscated"). 2 - it may be used in a future query (1.6 R2 behavior).
Thanks, Shawn.

From: Robert Schuff Sent: Thursday, February 03, 2011 3:23 PM To: i2b2 AUG Members
Subject: Re: Obfuscated user role....WAS: Re: user roles documentation?
OK. If the patient set is not usable by someone with just obfuscated role, does it make sense to show patient set as an option or result for them? There must be a use case I am not thinking of. I am however thinking about this from the training / user perspective and thinking about their reactions to being able to choose something they cannot use plus seeing that the number of patients returned is different. Seems like questions and suspicions are likely to abound.
Rob
Robert Schuff - Director, Clinical Research Informatics
Oregon Clinical & Translational Research Institute Oregon Health & Science University
http://www.octri.org/

From: Mike Mendis Date: Thu, 3 Feb 2011 10:12:57 -0800 To: Robert Schuff, i2b2 AUG Members
Subject: Re: Obfuscated user role....WAS: Re: user roles documentation?
Yes the obfuscated user is able to create a patient set. They are not able to do anything with it. If there role is increased than they will be able to use it and any previous patient set created.
That is fine that the count and set is different. The result being a obducated user will be off by ~3 by design.
If you run the same query mutliple times you will get different results, but you will also be locked out if you run it more than 7 times.

On 2/3/11 12:53 PM, "Robert Schuff" wrote:
Great! Thanks everyone.
OK so I am curious about the expected behavior for an obfuscated user (has roles USER and DATA_OBFSC) with respect to patient sets.
Should the obfuscated user be able to select patient set as a query return type?
Should the record counts for "Patient Count" and "Patient Set" match? In 1.5.1 VM they do not.
Should Patient Count be invariate across runs of a query for a user with only the DATA_OBFSC role?

Here is a screenshot shot apparent behavior of query results for a user with DATA_OBFSC role.
Note that the "Patient Count" appears to be invariant but the patient set does vary across runs. Is this the expected behavior?
Thanks!
Robert Schuff ?Director, Clinical Research Informatics
Oregon Clinical & Translational Research Institute Oregon Health & Science University
http://www.octri.org/

From: Arvinder Choudhary Date: Thu, 3 Feb 2011 08:17:04 ?0800 To: i2b2 AUG Members, Robert Schuff
Subject: Re: user roles documentation?
Rob:
Please find the documentation attached that I received from Mike Mendis.
Arvinder

Robert Schuff 2/3/2011 9:18 AM
Hi Folks,
Where can I find a listing and documentation of the expected behaviors for the various user roles in i2b2 such as Data_Obfuscation and De?identification role.
Thanks!
Rob
Robert Schuff ?Director, Clinical Research Informatics
Oregon Clinical & Translational Research Institute Oregon Health & Science University http://www.octri.org/


NEXT EMAIL
From: Mobed, Ketty Sent: Thursday, February 03, 2011 12:12 PM To: Brian Ostasiewski; i2b2 AUG Members
Subject: RE: Lab hierarchies in i2b2
My hope is that the current extensive LOINC code information (including labs) available in NCBO http://bioportal.bioontology.org/ontologies/44774 can be linked (and mapped) to the CPT ontology there (http://bioportal.bioontology.org/visualize/42994 ) which needs to be expanded to include all CPT codes and ultimately be used for i2b2 query purposes (using the Health Ontology Mapper).
Are there any thoughts on whether and when CPT codes might be freely available (and not as a proprietary ontology)?

From: Brian Ostasiewski Sent: Thursday, February 03, 2011 5:59 AM To: Mobed, Ketty; i2b2 AUG Members
Subject: RE: Lab hierarchies in i2b2
This is for sure one of the main problems we face, the coverage of the official mappings is not extensive enough. For now our users have to use the find feature to locate terms that are not in the CPT tree (and for which they do not know the LOINC number). The hope is that either a new mapping file gets released, or we get a physician resource here that can help us fill in the most used blanks.

From: Mobed, Ketty Sent: Monday, January 31, 2011 2:22 PM To: Brian Ostasiewski; i2b2 AUG Members
Subject: RE: Lab hierarchies in i2b2
As I am working on these LOINC to CPT lab value crosswalks, I also have noticed that not all CPT codes related to labs are contained in the imported RRF file.
Any thoughts?
Ketty

From: Brian Ostasiewski Sent: Thursday, January 27, 2011 11:11 AM To: i2b2 AUG Members
Subject: RE: Lab hierarchies in i2b2
I did it by hand. There's only about 2000 entries in MRSMAP.RRF text file (in the LNC215 _TO_CPT2005_MAPPINGS.zip file), which I imported into Excel - you can specify the pipe | as your column delimiter to read in the RRF files and then use the V/HLOOKUP functions to cross-reference to data in another spreadsheet, in this case using the NLM CUI codes (ex: C1704201) which are often a common key among files (detailed descriptions of files and fields here: http://www.ncbi.nlm.nih.gov/books/NBK9685/#ch03.I3313_Mappings_File_ )
There's a possibility I will write a script to do this in the future if I see us potentially regenerating the tree often, but unfortunately there is nothing at the moment.

From: Peter Beninato Sent: Thursday, January 27, 2011 1:02 PM To: Brian Ostasiewski; i2b2 AUG Members
Subject: RE: Lab hierarchies in i2b2
Hi Brian,
So I downloaded the zip and expanded it. The file suffix is .RRF.
Is that the umls methasarus format?
Could you outline the process to transform that download into the i2b2 ontology table data.
Thanks.
Peter

From: Brian Ostasiewski Sent: Wednesday, January 26, 2011 11:10 AM To: i2b2 AUG Members
Subject: RE: Lab hierarchies in i2b2
Ketty-The actual structure shown in my screencap is the Pathology and Lab branch of: http://bioportal.bioontology.org/visualize/42994
I believe the source of the CPT<->LOINC mappings was: http://www.nlm.nih.gov/research/umls/mapping_projects/loinc_to_cpt_map.html
Donglu Both trees reference the same underlying concept code (say "LOI|2436-4", so there is just one entry per in observation_fact).

From: Donglu Xie Sent: Wednesday, January 26, 2011 12:28 PM To: Brian Ostasiewski; Robert Schuff; i2b2 AUG Members
Subject: RE: Lab hierarchies in i2b2
Two trees are very easy and friendly for the end users to navigate.
How do you construct data in observation_fact and concept_dimension tables based on two trees?
Thanks.
Donglu

From: Mobed, Ketty Sent: Wednesday, January 26, 2011 12:25 PM To: Brian Ostasiewski; Robert Schuff; i2b2 AUG Members
Subject: RE: Lab hierarchies in i2b2
Thanks Brian!
Is this particular CPT tree available in NCBO BioPortal?
Ketty
Ketty Mobed, PhD MSPH Research Analyst / Terminologist ITS - Academic Research Systems UCSF - Campus Box 0707

From: Brian Ostasiewski Sent: Wednesday, January 26, 2011 9:06 AM To: Robert Schuff; i2b2 AUG Members
Subject: RE: Lab hierarchies in i2b2
We use two trees in the metadata – one solely by LOINC code number, in case the user knows the exact code, and another that is structured by CPT category (CPT concepts provide the tree folders, with many specific LOINC leafs under each CPT folder), this seems to be more intuitive to researchers.
Example:
Beyond LOINC we also have labs accessible via our internal structure by code and panel.

From: Robert Schuff Sent: Tuesday, January 25, 2011 3:02 PM To: i2b2 AUG Members
Subject: Lab hierarchies in i2b2
Hi Folks,
We at Oregon are investigating the best lab test hierarchy (not test name encoding..LOINC codes/terms at the leaf node) to use for building our search term tree in i2b2. Seems there really isn't a commonly used structure and out of the box i2b2 comes with the Partners-based hierarchy. At OHSU there trees coded within our EHR (Epic) but they are based on historical groupings also done locally. The LOINC RELMA tool seems to have some hierarchies in it but beyond that we have yet to identify a standard hierarchy.
So I am seeking advice from you all. My questions are:
How did you arrive at your selection for a hierarchy?
Is there an emerging standard taxonomy that can be used in i2b2 to build the search term tree?
Should we at least attempt then to create (or designate an existing) one as a "recommended" standard hierarchy.
Now certainly this does not even come close to addressing the myriad of issue surrounding labs (results) from an interoperability perspective, but may be one small important (and achievable?) step towards interoperability of i2b2 installations.
Thanks!
Rob
Robert Schuff - Director, Clinical Research Informatics
Oregon Clinical & Translational Research Institute Oregon Health & Science University http://www.octri.org/


NEXT EMAIL
From: Arvinder Choudhary Sent: Thursday, February 03, 2011 11:17 AM To: i2b2 AUG Members; Robert Schuff
Subject: Re: user roles documentation?
Rob:
Please find the documentation attached that I received from Mike Mendis.
Arvinder
Robert Schuff 2/3/2011 9:18 AM
Hi Folks, Where can I find a listing and documentation of the expected behaviors for the various user roles in
i2b2 such as Data_Obfuscation and De-identification role.
Thanks!
Rob Robert Schuff - Director, Clinical Research Informatics
Oregon Clinical & Translational Research Institute Oregon Health & Science University http://www.octri.org/


NEXT EMAIL
From: Niveditha Thota Sent: Tuesday, February 01, 2011 5:14 PM To: Phillips, Lori C.; Jeremy Nix Cc: i2b2 AUG Members; i2b2 install help at google groups
Subject: RE: Help with i2b2 installation
The port number for the PM cell was wrong in the workplace.properties file.
Changed it to 7070 and that took care of the error.
Thanks for your help.
Nivi

From: Phillips, Lori C. Sent: Tuesday, February 01, 2011 3:34 PM To: Niveditha Thota; Jeremy Nix Cc: i2b2 AUG Members; i2b2 install help at google groups
Subject: RE: Help with i2b2 installation
Not sure if this is your problem, but WORK_DB_LOOKUP would also have to be modified up in the same manner.
Is the address of the PM service configured correctly in the workplace properties file ?
workplace.ws.pm.url= ????
/axis2/rest/PMService/getServices — maybe should be /i2b2/rest/.....?

From: Niveditha Thota Sent: Tuesday, February 01, 2011 3:56 PM To: Jeremy Nix Cc: i2b2 AUG Members; i2b2 install help at google groups
Subject: RE: Help with i2b2 installation
Hi,
Thanks for looking into the issue. I followed your instruction and checked the C_DOMAIN_ID value for the CRC_DB_LOOKUP and ONT_DB_LOOKUP tables. Both the tables have value 'i2b2demo' instead of 'demo'. Instead of changing the database value, I modified the entry in the i2b2workbench.properties file to use the domain – i2b2demo. I also changed the domain name under environment variables under GLOBAL HIVE DATA in the gridsphere to i2b2demo. Doing this got rid of the the error - "Could not locate hive= demo".
The Workplace cell still has some problem. It gives error message – 'User was not validated'
JBOSS log shows:
15:03:33,107 ERROR [RESTSender] Error in processing POST request
org.apache.axis2.AxisFault: Transport error 404 . Error Message is <html><head><title> JBossWeb/2.0.1.GA - Error report</title><style><!-H1 {font-family:Tahoma,Arial,sansserif;color:white;background-color:#525D76;font-size:22px;} H2 {font-family:Tahoma,Arial,sansserif;color:white;background-color:#525D76;font-size:16px;} H3 {font-family:Tahoma,Arial,sansserif;color:white;background-color:#525D76;font-size:14px;} BODY {font-family:Tahoma,Arial,sansserif;color:black;background-color:white;} B {font-family:Tahoma,Arial,sansserif;color:white;background-color:#525D76;} P {font-family:Tahoma,Arial,sansserif;background:white;color:black;font-size:12px;}A {color : black;}A.name {color : black;}HR {color : #525D76;}-></style> </head><body><h1>HTTP Status 404 - /axis2/rest/PMService/getServices</h1><HR size="1" noshade="noshade"><p><b>type</b> Status report</p><p><b>message</b> <u>/axis2/rest/PMService/getServices</u></p><p><b> description</b> <u>The requested resource (/axis2/rest/PMService/getServices) is not available.</u></p><HR size="1" noshade="noshade"><h3>JBossWeb/2.0.1.GA</h3></body></html>
at org.apache.axis2.transport.http.RESTSender.sendViaPost(RESTSender.java:295)
at org.apache.axis2.transport.http.RESTSender.send(RESTSender.java:141)
at org.apache.axis2.transport.http.CommonsHTTPTransportSender.writeMessageWithCommons(Com monsHTTPTransportSender.java:334)
at org.apache.axis2.transport.http.CommonsHTTPTransportSender.invoke(CommonsHTTPTransportSe nder.java:205)
at org.apache.axis2.engine.AxisEngine.send(AxisEngine.java:670)
at org.apache.axis2.description.OutInAxisOperationClient.send(OutInAxisOperation.java:365) at org.apache.axis2.description.OutInAxisOperationClient.execute(OutInAxisOperation.java:295) at org.apache.axis2.client.ServiceClient.sendReceive(ServiceClient.java:579) at org.apache.axis2.client.ServiceClient.sendReceive(ServiceClient.java:508) at edu.harvard.i2b2.pm.ws.PMServiceDriver.sendREST(Unknown Source) at edu.harvard.i2b2.pm.ws.PMServiceDriver.getRoles(Unknown Source) at edu.harvard.i2b2.workplace.delegate.RequestHandler.getRoleInfo(Unknown Source) at edu.harvard.i2b2.workplace.delegate.GetFoldersByProjectHandler.<init>(Unknown Source) at edu.harvard.i2b2.workplace.ws.WorkplaceService.getFoldersByProject(Unknown Source) at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method) at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:39)
at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:25) at java.lang.reflect.Method.invoke(Method.java:585) at
org.apache.axis2.receivers.RawXMLINOutMessageReceiver.invokeBusinessLogic(RawXMLINOutMe ssageReceiver.java:88)
at org.apache.axis2.receivers.AbstractInOutSyncMessageReceiver.receive(AbstractInOutSyncMessage Receiver.java:39)
at org.apache.axis2.engine.AxisEngine.receive(AxisEngine.java:493) at org.apache.axis2.transport.http.util.RESTUtil.invokeAxisEngine(RESTUtil.java:153) at org.apache.axis2.transport.http.util.RESTUtil.processPostRequest(RESTUtil.java:102) at org.apache.axis2.transport.http.AxisRESTServlet.doPost(AxisRESTServlet.java:66) at javax.servlet.http.HttpServlet.service(HttpServlet.java:710) at javax.servlet.http.HttpServlet.service(HttpServlet.java:803) at
org.apache.catalina.core.ApplicationFilterChain.internalDoFilter(ApplicationFilterChain.java:290) at org.apache.catalina.core.ApplicationFilterChain.doFilter(ApplicationFilterChain.java:206)
at org.jboss.web.tomcat.filters.ReplyHeaderFilter.doFilter(ReplyHeaderFilter.java:96)
at org.apache.catalina.core.ApplicationFilterChain.internalDoFilter(ApplicationFilterChain.java:235)
at org.apache.catalina.core.ApplicationFilterChain.doFilter(ApplicationFilterChain.java:206)
at org.apache.catalina.core.StandardWrapperValve.invoke(StandardWrapperValve.java:230)
at org.apache.catalina.core.StandardContextValve.invoke(StandardContextValve.java:175)
at org.jboss.web.tomcat.security.SecurityAssociationValve.invoke(SecurityAssociationValve.java:179)
at org.jboss.web.tomcat.security.JaccContextValve.invoke(JaccContextValve.java:84)
at org.apache.catalina.core.StandardHostValve.invoke(StandardHostValve.java:127)
at org.apache.catalina.valves.ErrorReportValve.invoke(ErrorReportValve.java:102)
at org.jboss.web.tomcat.service.jca.CachedConnectionValve.invoke(CachedConnectionValve.java:157)
at org.apache.catalina.core.StandardEngineValve.invoke(StandardEngineValve.java:109)
at org.apache.catalina.connector.CoyoteAdapter.service(CoyoteAdapter.java:262)
at org.apache.coyote.http11.Http11Processor.process(Http11Processor.java:844)
at org.apache.coyote.http11.Http11Protocol $Http11ConnectionHandler.process(Http11Protocol.java:583)
at org.apache.tomcat.util.net.JIoEndpoint$Worker.run(JIoEndpoint.java:446)
at java.lang.Thread.run(Thread.java:595) 15:03:33,108 ERROR [PMServiceDriver] Incoming message input stream is null 15:03:33,108 ERROR [GetFoldersByProjectHandler] Cant connect to PM service

Nivi

From: Jeremy Nix Sent: Tuesday, February 01, 2011 1:36 PM To: Niveditha Thota Cc: i2b2 AUG Members; i2b2 install help at google groups
Subject: Re: Help with i2b2 installation
The message of importance here is "Could not locate hive= demo". This error is thrown in both the Ontology Cell and the CRC cell if the cell is unable to read out the proper datasource information from the crc_db_lookup or ont_db_lookup table in the i2b2hive schema. Verify the entries in these tables to make sure that there are entries with a C_DOMAIN_ID of "demo". If none exist, then that's at least a starting point. If the records do exist, then you will want to verify the contents of the following files:
jboss/server/default/conf/crcapp/CRCApplicationContext.xml
make sure that the datasource entry for CRCDataSourceLookup is using the same i2b2hive schema/user/database that you used above to verify the contents of the crc_db_lookup table.
jboss/server/deploy/ont-ds.xml
make sure that the datasource entry for OntologyBootStrapDS is using the same i2b2hive schema/user/database that you used above to verify the contents of the ont_db_lookup table.
Hope this helps
Jeremy Nix Senior Application Developer Cincinnati Children's Hospital Medical Center

On 02/01/2011 01:37 PM, Niveditha Thota wrote:
Hi,
I am working with i2b2 version 1.3 VM here at UAB. The VM is installed in UNIX environment. We are trying to point the cells in i2b2 to our production database which is located on a remote machine. To achieve this, I have modified the relevant database files and recompiled the source files for the cells following the directions from the installation guide. But, we are facing trouble getting the workbench to work with the i2b2 instance configured for the production database.
Three of the cells in the workbench are giving the following errors:
Ontology cell is unavailable
Workplace cell is unavailable
Data Repository cell is unavailable
The logs from the JBOSS have the error below:
13:12:19,830 ERROR [AxisRESTServlet] org.apache.axis2.AxisFault: Could not locate hive= demo; nested exception is:
edu.harvard.i2b2.common.exception.I2B2Exception: Could not locate hive= demo; nested exception is:
org.apache.axis2.AxisFault: Could not locate hive= demo; nested exception is:
edu.harvard.i2b2.common.exception.I2B2Exception: Could not locate hive= demo
13:12:21,494 INFO [PropertiesFactoryBean] Loading properties file from class path resource [crc_application_directory.properties]
13:12:21,754 INFO [XmlBeanDefinitionReader] Loading XML bean definitions from URL [[file:/opt/jboss/server/default/conf/crcapp/CRCApplicationContext.xml]]
13:12:22,164 INFO [CollectionFactory] JDK 1.4+ collections available
13:12:22,192 INFO [CollectionFactory] Commons Collections 3.x available
13:12:22,195 INFO [FileSystemXmlApplicationContext] Bean factory for application context [org.springframework.context.support.FileSystemXmlApplicationContext;hashCode=15850979]: org.springframework.beans.factory.support.DefaultListableBeanFactory defining beans [jaxbPackage,appType,message_header,TestDataSourceLookup,TestDataSource,CRCDataSourceL ookup,defaultSetfinderResultType,setFinderResultGeneratorMap]; root of BeanFactory hierarchy
13:12:22,208 INFO [FileSystemXmlApplicationContext] 8 beans defined in application context [org.springframework.context.support.FileSystemXmlApplicationContext;hashCode=15850979]
13:12:22,214 INFO [FileSystemXmlApplicationContext] Unable to locate MessageSource with name 'messageSource': using default [org.springframework.context.support.DelegatingMessageSource@b05409]
13:12:22,218 INFO [FileSystemXmlApplicationContext] Unable to locate ApplicationEventMulticaster with name 'applicationEventMulticaster': using default [org.springframework.context.event.SimpleApplicationEventMulticaster@13dff89]
13:12:22,219 INFO [DefaultListableBeanFactory] Pre-instantiating singletons in factory [org.springframework.beans.factory.support.DefaultListableBeanFactory defining beans [jaxbPackage,appType,message_header,TestDataSourceLookup,TestDataSource,CRCDataSourceL ookup,defaultSetfinderResultType,setFinderResultGeneratorMap]; root of BeanFactory hierarchy]
13:12:23,429 INFO [PropertiesFactoryBean] Loading properties file from class path resource [crc_application_directory.properties]
13:12:23,429 INFO [PropertiesFactoryBean] Loading properties file from file [/opt/jboss/server/default/conf/crcapp/crc.properties]
13:12:23,592 INFO [QueryRequestDelegate] project name from PM Demo Group
13:12:23,593 INFO [QueryRequestDelegate] project id from PM Demo 13:12:23,593 ERROR [QueryRequestDelegate] Project role not set for the user
13:12:23,762 INFO [DataSourceLookupHelper] Could not match Project id=[/Demo/] Trying with hive =demo
13:12:24,239 INFO [PropertiesFactoryBean] Loading properties file from class path resource [workplace_application_directory.properties]
13:12:24,241 INFO [PropertiesFactoryBean] Loading properties file from file [/opt/jboss/server/default/conf/workplaceapp/workplace.properties]
13:12:24,262 ERROR [RESTSender] Error in processing POST request
org.apache.axis2.AxisFault: Transport error 404 . Error Message is <html><head><title> JBossWeb/2.0.1.GA - Error report</title><style><!-H1 {font-family:Tahoma,Arial,sansserif;color:white;background-color:#525D76;font-size:22px;} H2 {font-family:Tahoma,Arial,sansserif;color:white;background-color:#525D76;font-size:16px;} H3 {font-family:Tahoma,Arial,sansserif;color:white;background-color:#525D76;font-size:14px;} BODY {font-family:Tahoma,Arial,sansserif;color:black;background-color:white;} B {font-family:Tahoma,Arial,sansserif;color:white;background-color:#525D76;} P {font-family:Tahoma,Arial,sansserif;background:white;color:black;font-size:12px;}A {color : black;}A.name {color : black;}HR {color : #525D76;}-></style> </head><body><h1>HTTP Status 404 - /axis2/rest/PMService/getServices</h1><HR size="1" noshade="noshade"><p><b>type</b> Status report</p><p><b>message</b> <u>/axis2/rest/PMService/getServices</u></p><p><b> description</b> <u>The requested resource (/axis2/rest/PMService/getServices) is not available.</u></p><HR size="1" noshade="noshade"><h3>JBossWeb/2.0.1.GA</h3></body></html>
at org.apache.axis2.transport.http.RESTSender.sendViaPost(RESTSender.java:295)
at org.apache.axis2.transport.http.RESTSender.send(RESTSender.java:141)
at org.apache.axis2.transport.http.CommonsHTTPTransportSender.writeMessageWithCommons(Com monsHTTPTransportSender.java:334)
at org.apache.axis2.transport.http.CommonsHTTPTransportSender.invoke(CommonsHTTPTransportSe nder.java:205)
at org.apache.axis2.engine.AxisEngine.send(AxisEngine.java:670)
at org.apache.axis2.description.OutInAxisOperationClient.send(OutInAxisOperation.java:365)
at org.apache.axis2.description.OutInAxisOperationClient.execute(OutInAxisOperation.java:295)
at org.apache.axis2.client.ServiceClient.sendReceive(ServiceClient.java:579)
at org.apache.axis2.client.ServiceClient.sendReceive(ServiceClient.java:508)
at edu.harvard.i2b2.pm.ws.PMServiceDriver.sendREST(Unknown Source)
at edu.harvard.i2b2.pm.ws.PMServiceDriver.getRoles(Unknown Source)
at edu.harvard.i2b2.workplace.delegate.RequestHandler.getRoleInfo(Unknown Source) at edu.harvard.i2b2.workplace.delegate.GetFoldersByProjectHandler.<init>(Unknown Source) at edu.harvard.i2b2.workplace.ws.WorkplaceService.getFoldersByProject(Unknown Source) at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method) at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:39) at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:25) at java.lang.reflect.Method.invoke(Method.java:585) at
org.apache.axis2.receivers.RawXMLINOutMessageReceiver.invokeBusinessLogic(RawXMLINOutMe ssageReceiver.java:88)
at org.apache.axis2.receivers.AbstractInOutSyncMessageReceiver.receive(AbstractInOutSyncMessage Receiver.java:39)
at org.apache.axis2.engine.AxisEngine.receive(AxisEngine.java:493) at org.apache.axis2.transport.http.util.RESTUtil.invokeAxisEngine(RESTUtil.java:153) at org.apache.axis2.transport.http.util.RESTUtil.processPostRequest(RESTUtil.java:102) at org.apache.axis2.transport.http.AxisRESTServlet.doPost(AxisRESTServlet.java:66) at javax.servlet.http.HttpServlet.service(HttpServlet.java:710) at javax.servlet.http.HttpServlet.service(HttpServlet.java:803) at
org.apache.catalina.core.ApplicationFilterChain.internalDoFilter(ApplicationFilterChain.java:290) at org.apache.catalina.core.ApplicationFilterChain.doFilter(ApplicationFilterChain.java:206) at org.jboss.web.tomcat.filters.ReplyHeaderFilter.doFilter(ReplyHeaderFilter.java:96) at
org.apache.catalina.core.ApplicationFilterChain.internalDoFilter(ApplicationFilterChain.java:235) at org.apache.catalina.core.ApplicationFilterChain.doFilter(ApplicationFilterChain.java:206) at org.apache.catalina.core.StandardWrapperValve.invoke(StandardWrapperValve.java:230) at org.apache.catalina.core.StandardContextValve.invoke(StandardContextValve.java:175) at
org.jboss.web.tomcat.security.SecurityAssociationValve.invoke(SecurityAssociationValve.java:179)
at org.jboss.web.tomcat.security.JaccContextValve.invoke(JaccContextValve.java:84) at org.apache.catalina.core.StandardHostValve.invoke(StandardHostValve.java:127) at org.apache.catalina.valves.ErrorReportValve.invoke(ErrorReportValve.java:102) at
org.jboss.web.tomcat.service.jca.CachedConnectionValve.invoke(CachedConnectionValve.java:157) at org.apache.catalina.core.StandardEngineValve.invoke(StandardEngineValve.java:109) at org.apache.catalina.connector.CoyoteAdapter.service(CoyoteAdapter.java:262) at org.apache.coyote.http11.Http11Processor.process(Http11Processor.java:844) at org.apache.coyote.http11.Http11Protocol
$Http11ConnectionHandler.process(Http11Protocol.java:583) at org.apache.tomcat.util.net.JIoEndpoint$Worker.run(JIoEndpoint.java:446) at java.lang.Thread.run(Thread.java:595)
13:12:24,263 ERROR [PMServiceDriver] Incoming message input stream is null 13:12:24,263 ERROR [GetFoldersByProjectHandler] Cant connect to PM service 13:12:24,322 ERROR [AxisRESTServlet] org.apache.axis2.AxisFault: Could not locate hive= demo;
nested exception is: edu.harvard.i2b2.common.exception.I2B2Exception: Could not locate hive= demo; nested
exception is: org.apache.axis2.AxisFault: Could not locate hive= demo; nested exception is: edu.harvard.i2b2.common.exception.I2B2Exception: Could not locate hive= demo

Any thoughts on what might be wrong?
Thanks, Nive


NEXT EMAIL
From: Jeff Cowall Sent: Tuesday, February 01, 2011 11:11 AM To: i2b2 AUG Members
Subject: How do Patient Dimensions across multiple Project/(Oracle)Schemas relate in a given hive?
I think I understand the Observation_Fact - Concept_Dim - encounter- provider relationships across multiple projects/(Oracle)schemas.
And how a project relates to a path in the ontology cell.
But can I get away with one master Patient Dimension in a common project/schema?
Am I missing something?
Jeff, trying to plan ahead.
Jeff Cowall Sr Data Architect University of Michigan Health System MCIT Clinical Research