Project Title: Comprehensive Health, Environment, Safety System (CHESS)

Sautter Award Nomination - CHESS- Lawrence Berkeley National Laboratory

Project Title: Comprehensive Health, Environment, Safety System (CHESS)

Submitter: GitaMeckel

Project Team:

EHSBusiness Analysts: Lee Aleksich, Maria Prieto, Randy Roig

IT Personnel: Vijay Amirineni, Vanessa Arce-Kaskiri, Lana Newman, Jeremy Pease, Steve Wyrick

Key EHS Subject Matter Experts: Susan Broadway, Ross Fisher, Peter Lichty, Tim Roberts, Herb Toor

Summary:

The Environment, Health and Safety (EHS) strategic information management modernization program of the Lawrence Berkeley National Laboratory (LBNL) is a five-year project to modernize the wide range of information management functions that the EHS Divisionand the LBNL community depends upon and which had been supported by approximately 40+ different legacy systems, most of which were one-off custom applications 5-15 years old. The core approach was to replace these systems by consolidating a broad range of technical ES&H business functions into a small number of state-of-the-art COTS applications. This goal was to yield superior end-to-end business process integration, improved customer service, lower operational costs, superior management information, and overall improved efficiency, completeness, and quality of EHS business operations.

The first major phase of this modernization effort was the acquisition and implementation of the SAI Global Cintellateplatform that replaced approximately multiple legacy applications in the health and safety programs: Industrial Hygiene (IH), Injury/Illness & Absence Management, Incident Investigation & Reporting, and Medical Surveillance into one core enterprise system called CHESS. Tracking of the improvements in operational efficiency and effectiveness has demonstrated substantial gains achieved through this initiative over the last two years. This successful innovation will serve as a model for future improvements in EHS IT projects.

Project Description:

The origin of this project was a 2010 ExecutiveReview of existingIT infrastructuresupportingthe EHS Division. The goalswere to develop a strategic roadmap to establish a modern and robust technology environment with an emphasis on (1) improving operational efficiencies and effectiveness; (2) improving service and EHS data management delivery to customers and stakeholders; and (3) improving regulatory compliance. The review covered the whole range of EHS IT systems in use in the EHS Division at that time, totaling 40+ stand-alone applications. EHS software applications were developed over many years using disparate technology, often in urgent response to DOE drivers/commitments, and usually as stand-alone applications. Most of the applications were custom developed at LBNL and in stages due to limited funding and have been incrementally modified over the years to meet additional user requirements, regulatory mandates, and other needs. The result of this approach is that the systems had become needlessly complex and disjointed.

The other fundamental shift used in the development of EHS systems was to follow industry standard Software Development LifeCycle (SDLC) processes based on elements of the Capability Maturity Model Implementation and Six Sigma.

The health and safety functions were selected to be the first to be modernized. LBNL’s EHS business analystperformed an end-to-end business process analysis. Each business process was examined in its “as is” form, evaluated, and, where necessary, re-engineered to support improved operational efficiency and effectiveness, more complete regulatory compliance, and better process flow. Each review involved an interdisciplinary team of business analysts, subject matter experts (SMEs), customers, and stakeholders. The teams employed a variation of a Lean Six Sigma DMAIC process to gather and assess the current business processes and to recommend improvements. These changes were expected to drive performance and reduce costs byproviding improved management reporting, ensuring improved workflow, and, where cost effective, automating steps currently done manually. Once the business process re-engineering was complete, the EHS business analyst team determined how the SAI Global Cintellateplatform could be configured to support the new process, and developed the necessary design requirements and specifications.

The firstimplementation phasefocused on Medical Surveillance and Industrial Hygiene (IH). This involved replacing all or part of the following applications with the SAI Global Cintellateplatform: CTS (for IH Health and Safety Exposure Hazard Assessment), OHM (medical surveillance), Bloodborne Pathogens, Respiratory Referrals, and Respirator Fit Tests. This work was completed and the system was rolled out in December 2011 for IH and in April 2012 for Medical Surveillance/Bloodborne Pathogens/Respiratory Protection.

The second implementation phase started in December 2011 and covered the remaining types of medical exams provided by the medical clinic, the Injury/Illness & Absence Management Program and Incident Investigation & Reporting Program (SAAR). The implementation of these programs is now underway and is expected to be rolled out by early June 2013.

The goals of the effortwere to achieve the following improvements:

• Greatly improve work processes, yielding consistent results of much higher quality in a shorter period of work effort and/or elapsed time;
• Replace and retire many obsolete IT systems that required significant maintenance and were no longer supported by the vendor;
• Reduce data quality problems and improve handoffs between departments, since data and processes must be shared and linked within a single unified application platform; and
• Improve metrics and reporting that allow management to understand and control business processes.

In this submittal, we have chosen to summarize the problems and outcomes of the project for Industrial Hygiene, Injury/Illness and Absence Management, and Incident Investigation & Reporting. We elected to use the Medical Surveillance part of the CHESS system to provide a more in depth analysis of the problems facing the program as well as the benefits and improvements in the new system.

1)Industrial Hygiene

Industrial Hygiene is the measurement and evaluation of employee exposure to potential chemicals or toxins. The process starts with a request or identification of a need for an exposure survey, proceeds through the data collection and analysis, and ends with the publishing of a report and applying that report to workplace controls. Based on the measured exposure, the work might need to be redesigned, have added controls, or be reconsidered.

Problems:Prior to CHESS, the Industrial Hygiene program had no formal organized and IT-supported work flow process – the request queue was an excel spreadsheet, there was no fully standardized method of survey documentation, and except for very high profile issues, there was only an informal and mostly undocumented peer review process. The CTS computer application itself was obsolete and was no longer supported by the vendor. It was also difficult to use (the Lab had two “data entry specialists” as most industrial hygienists found the program unusable)and almost impossible to query to find and use previous IH surveys.

Outcomes: With the advent of CHESS and the redesign of the business process, survey requests proceed from a formal prioritized study queue through a series of options of study mode to the survey itself. The survey results are fully documented and peer reviewed within an automated system. Most importantly, if a previous survey is directly applicable to a new situation, it can easily be found and referenced, allowing it to be used as a justification to avoid the conduct of a duplicate study or to significantly reduce the number ofexposure samples that are needed in the new study.

2)Injury/Illness and Absence Management

Injury/Illness and Absence management involves the clinic response to occupational injuries and illnesses, workers compensation, family and medical leave, and return to work. As a DOE facility, the Lab has substantive additional requirements for these processes. For example, every person absent from work for more than 5 days due to illness must bring a doctor’s note and/or be reviewed by the clinic for “fitness for duty” prior to returning to work.

Problems: Prior to CHESS, the Injury & Illness program and absence management program within the clinic used OHM, a program purchased in the mid-1990s along with spreadsheets and other hard copy tracking methods. Entry into OHM is difficult because of the user interface. All communication was done either by phone call or e-mail. Further, whenever an occupational illness or injury occurred, that information had to be hand-transferred into another computer system (SAAR) for incident investigation and reporting. The overall business process for this topic was slow, labor intensive, and prone to errors.

Outcomes: With the advent of CHESS and the redesign of the business processes, there is now a smooth flow of information gathering into the various reporting and tracking process, grouping together information related to a “case” in a manner that allows quick query and reference to the previous history. For occupational cases, there is a smooth and automated handoff of data to the Incident Investigation and Reporting process.

3)Incident Investigation & Reporting

All OSHA and DOE recordable injuries and illness must be reported to the regulators and investigated as to their cause. The results of the investigation must be evaluated to determine if work practices or physical structures need to be altered to prevent recurrence.

Problems: As noted above, prior to CHESS, the incident investigation system depended on a copy-and-paste transfer of information from OHM into SAAR. The investigation itself was mostly done outside any computer system and then uploaded at the end. Since LBNL is subject to DOE regulations in addition to OSHA, two different sets of “injury books” needed to be kept. Most of the regulatory recordkeepingfor DOE and OSHA was completedby manually transferring informationfrom OHM to Excel, correcting it to reflect the two sets of requirements, and storing it in a set of “archive” spreadsheets manually updated by the Subject Matter Expert.

Outcomes: With the introduction of CHESS, the process of information transfer, investigative work, reporting and approval is done within the CHESS system (and much of it is automated). For example, most of the gathering of accident and work hour statistics is now automated, reducing the probability of error and freeing up time for the SME. Further, the new investigative approach is flexible, allowing the team and investigation to be customized to the particular event.

Detailed Benefits for Medical Surveillance

Both OSHA and DOE require baseline and periodic medical surveillance exams for employees who work with certain types of chemicals or have specified jobs. The process starts with a subject matter expert (SME) who identifies how and if a particular regulation applies to LBNL operations and then identifies the individuals that require the surveillance. The individuals must then be “enrolled” in a medical surveillance group and provided the appropriate medical exams, either by the LBNL’s onsite clinic or by an outside provider. Until the employee has the surveillance completed or renewed, they are not authorized to do any work covered by the specific surveillance program.

Problems: Prior to the use of CHESS, the medical surveillance program consisted of four disconnected IT programs (OHM, BBP, Respiratory Referral, and Fit Test) and two significant manualdata-transfer processes that were highly prone to error. Below are a few examples of the problems found during the evaluation of the existing processes:

• Overall enrollment accuracy was only about 65%. In practice, the method developed by the SME to identify employees who needed surveillancewas either difficult to apply or not fully followed. As a result, some employees were surveyed who did not need it and some who required surveillance were overlooked. SMEs had no access to the list of employees who were enrolled and thus could not verify that employees were correctly identified.
• The main computer program used for medical surveillance, OHM, did not have detailed search and/or reporting capabilities. For example, the clinic could not project the upcoming exam backlog more than 2 weeks out. Similarly, nurses were unable to tell if a person had taken a similar exam in the past except by hand comparison of written medical records. As a result, tests were often duplicated.
• The OHM system itself was extremely difficult to use and was thus prone to errors – about 35% of all clinic visits had one or more data errors and 10% had more than 4 data errors per visit.
• There was no easy method to tell which clinic visits included lab resultssuch as blood tests, which are sent to external laboratories for analysis. Thus, lab results coming back via mail or fax tended to stack up and be entered into the system only when time was available. As a result, the average time to close a clinic visit file was over 16 days and 20% of the visits were open longer than 30 days. This means that credit was not granted for the surveillance during this time period and people were often delayed in obtaining or renewing internal licenses (such as for forklift drivers).
• The scheduling process for clinic examination rooms and health care personnel was inefficient and ineffective, meaning that the patient throughput in the clinic was limited. There was no method to catch up with a backlog except by hiring extra health care professionals on a temporary basis.
• All communication with patients and staff was hand done either via email or telephone by the receptionist and the professional staff. This took a significant amount of time away from patient care.

Outcomes: CHESS’s medical surveillance module worked to address all of these issues.

• 90% of all enrollments of employees into medical surveillance groups are automated with a high (>99%) degree of accuracy. This figure is expected to rise to 96% over the next year while maintaining the same degree of accuracy.
• The system automatically produces an “exam due” and “exam scheduled” backlog report every Monday, allowing the clinic to plan their workload.
• The scheduling process was changed to specifically target room use by type of exam and to schedule the rooms more completely. As a result, the clinic now processes about two to three times the number of people per week over the prior approach with the same number of staff. This increased capacity helped reduce the exam backlog from about 1650 exams prior to CHESS to about 390 exams now despite a 10% increase in the number of people in the surveillance programs (refer to Medical Exams Due and Scheduled graph).
• Exam offers and appointment reminders are now automated. When the medical surveillance is voluntary and can be declined, declinations can be recorded electronically, without need to submit a hard copy declination.
• By using mistake-proofing techniques in the design of the data entry process, the number of clinic visits having data errors has dropped dramatically, from about 35% to about 2% (see Comparison of Error Rate in Medical Exam Entry graph).
• After talking to customers, we determined that 95% of all clinic visits should be completed within 10 days (including exam credit documentation). Meeting this goal required two implementation steps, as is shown in the Control Chart of Clinic Visit Completion Time by Stagegraph. Initially, although the average time was reduced from 16 to 9 days, 25% took more than 10 days. We addressed that issue by (a) making it easier in the program to close the visit; (b) improving the visibility of open visits on a calendar (see screen shot) and (c) providing a quick link to the medical record from the calendar to allow the nurse to access the file and enter the data needed to complete the visit. The average time to close a visit was 2.1 days for Q1 of 2013.
• We are just implementing another major improvement: automatic cross crediting of similar medical exams. In the past, similar exams were often repeated since verifying their past completion and giving credit required looking in hard copy records. CHESS initially helped this process by making the past history electronically available to the person conducting the exam, but the crediting process is manual and has been usually applied only when the person is already at the clinic. As of June 2013, CHESS will now look at all of the previous exams a person has taken as a new exam becomes due and, based on a rules-based process, will automatically give credit. Credit will not only be granted for the same exact type of exam (e.g. medical history for medical history), but also for “lesser” but different exams (e.g., crediting a whisper hearing test if the person has had a recent audiometer exam). We expect to avoid about 500 to 800 exams per year, a significant savings both in the time for medical personnel and in time spent away from work by employees.

Conclusion

The methods used in developing and implementing CHESS are a very effective model for gaining substantive improvements in operational efficiency and effectiveness in the EHS Division at LBNL. We believe that fully understanding and improving the business processes prior to developing an IT system and involving an interdisciplinary team of business analysts, subject matter experts (SMEs), customers, and stakeholders throughout the life of the project will yield superior long-term results. We anticipate that SAI Global Cintellatewill serve asacore enterprise system across ES&H technical domains for future system improvements and implementations within the EHS Division.

1