How Is Your Biobank Handling Disaster Recovery Efforts?

Fisher Bioservices (formerly McKesson BioServices) operates 15 separate sites in the Rockville, Maryland area with biorepositories storing over 30 million samples and drug repositories holding over $10 million worth of experimental drugs and reagents (many unique and irreplaceable) used for cancer and AIDS research. In September, 2003, the East Coast of the United States was struck by Hurricane Isabel, a category 2 storm as it struck the North Carolina coast moving northward. Over 6 million people across the mid-Atlantic region were without power, many for several weeks. In Maryland alone, 1.24 million people lost power. The BioServices business unit had previously experienced power outages over its 15 year history, but none exceeding 24 hours up until that time. During and following Hurricane Isabel, commercial power was out for up to 5 days for some of the BioServices sites. In addition, for a period of approximately 24 hours, all travel the area was halted by a government executive order.

Being that the event was a hurricane, the Maryland team had several days to plan and prepare. BioServices was well equipped with emergency generators as well as having many other systems in place in the event of power outages and disasters, and it had a well refined (and tested) emergency response plan in place. In the days prior to the hurricane reaching the area, all plans were reviewed and preparations made. All sites were equipped with generators and each one was checked and tested (Note that this is done on a routine basis, but the efforts were repeated in anticipation of the hurricane). Diesel tanks were topped off. Liquid nitrogen freezers were filled, as were the bulk tanks. Personnel were assigned to be on-site 24 hours a day until the event had cleared the area.

The Result

Overall the emergency plan functioned as designed. No samples or products were lost or endangered. Movement of materials in and out of the repositories was reestablished as soon as commercial transport was restored. The staff was extremely cooperative and more than willing to do what was necessary to assure that all systems were maintained in the best possible fashion. Following the hurricane, the BioServices management team conducted a thorough debriefing to review everything that happened. The following is a brief synopsis of what we learned as a result of this event.

What Went Right

The emergency plan was implemented, and for the most part was very effective. While power was lost, all but one of the 22 generators started automatically. The one that did not start immediately was brought on line with a few hours. While the generators did not provide 100% of the facility capacity needs, freezers and all critical equipment were covered. In addition, for sites that did not have sufficient emergency power to operate the HVAC systems to full capacity, large industrial fans were brought in to maintain air circulation for the freezers in order to minimize overheating of the freezers and other equipment.

A “hot site” at the company headquarters was also in place, and this site was fully covered by diesel power generators. This site provided landline telephone coverage, computer support including servers, and internet access so that customers and staff could be kept abreast of events. Key operations and maintenance staff were equipped with “push to talk” cell phones so the cell phone blockage that has occurred in other situations (example, the recent Boston marathon tragedy) was not an issue.

What Went Wrong

The generators were fueled by either diesel or natural gas. The natural gas supply was not interrupted, so no issues occurred here. Most of the diesel generators have “belly tanks” that provide fuel for 24–48 hours. Since this period was exceeded, we had to find fuel to keep them going. What we had not done prior to the event was to interface with our fuel vendors to ensure that we were on the critical list to be resupplied. Some serious arm twisting and begging was required to assure deliveries. At one point staff was sent out with 5-gallon containers (which had been purchased as part of the preparation procedures) to seek diesel from service stations and similar sources. This was a problem because many of the service stations either had no electricity to run their fuel pumps or their fuel supplies had been exhausted.

What Was Changed To Mitigate Future Disasters

Supply agreements were established with the fuel suppliers to get priority deliveries in the event of future disasters. We also went to the electric power suppliers and asked to be placed as high as possible on the restoration list (behind hospitals and life support services). We looked into getting our own diesel bulk tank(s) but that proved not to be feasible due to economic and environmental reasons. Most new generators were specified to be fueled by natural gas rather than diesel. Where economically viable, new generators were purchased as “duel fuel” units. These generators can be powered either by natural gas and propane or by diesel and propane (at a reduce power output) providing greater flexibility in emergency situations. Because generators cannot start or fail in service, in new locations we began to install pre-wired connectors to allow trailer mounted mobile generators to be brought in and rapidly connected – think of a giant extension cord. This is a relatively low cost modification or addition.

SOPs were also reviewed and staff training was augmented in light of the lessons learned as a result of this event, to maximize preparedness for future events.

Address correspondence to:

Phil Baird *

Principle

Daingerfield Consulting

E-mail: philmbaird@gmail.com

In 1978, the Harvard Brain Tissue Resource Center (HBTRC) was established at McLean Hospital by Edward D. Bird, MD, to provide high-quality postmortem brain tissue for the study of neurodegenerative and neuropsychiatric disorders. The total number of cases processed each year is approximately 275. The HBTRC receives these brain donations through community-based sources distributed throughout the country. Ninety percent of HBTRC tissue goes to investigators throughout the United States, while the remaining 10% is distributed to neuroscientists in Canada, Europe, the Middle East, Asia, South America, and Australia. On an annual basis, the HBTRC distributes several thousand tissue samples to 50–60 investigators. At any given time, the HBTRC has approximately 3000 postmortem brains stored in 26–80°C freezers. The HBTRC staff work diligently to avoid freezer breakdowns that will put its tissue collections in great jeopardy.

Freezer Failure

On June 1, 2012, the HBTRC experienced a catastrophic freezer failure that affected approximately 139 postmortem brains; they were destroyed by warming of the tissue. The cases stored in “Freezer U” represented a broad range of diagnoses, although the autism brain collection of Autism Speaks had the highest density of cases affected (n=52). During the days prior to the discovery of the failure, this freezer consistently registered -79°C, and the lights suggested that various aspects of the freezer were functioning appropriately. The summary readout on the front panel indicated that the “System is OK.” However, as we were soon to discover, these indicators belied the fact that the ambient temperature within the freezer had warmed to +7°C. We estimate that the warming pattern that resulted in this full-scale failure must have begun approximately 2–3 days before its discovery. It is important to note that our staff had been monitoring the freezers on a regular basis as they prepared brains for storage or distribution. At the time of the failure, all of our freezers had two parallel alarm systems: one that is intrinsic to each unit and triggers an alarm in the hospital security department, and another that is redundant and independently contacts HBTRC staff members if the internal temperature of a freezer begins to fall below -70°C. Neither of these alarms had been triggered during the period that Freezer U was failing. The failure of Freezer U was discovered when a staff member opened it to remove tissue for distribution purposes. As he did so, it was clear that the tissue had thawed and become very soft. The tissue in Freezer U was immediately transferred to a backup freezer equilibrated to -80°C. The facilities department at the hospital was immediately notified, and they cordoned off Freezer U so that a careful investigation to determine the cause of the failure could be launched.

A variety of technicians, engineers, and scientists who were quite familiar with standard research units evaluated Freezer U but indicated that they had never seen such a dramatic disparity between the temperature within the freezer and that registered on the front panel. A consulting engineer noted that the signals coming from the front panel and those associated with the main alarm converged at a single point on the central circuit board for this unit. The parallel alarm system operated independently of this freezer circuitry and was wired into a separate control system through the telephone cable within the room. A formal investigation was conducted by a private electronics engineering company that had dealt with many such situations at other institutions. This investigation concluded that all of the components of the freezer were not defective and concluded that it was not possible to ascertain why Freezer U failed.

What is particularly perplexing about this incident is the fact that the redundant alarm system, which had a thermal sensor inserted into the inner chamber of the freezer, also failed. Based on this observation, there was speculation in the news media that sabotage could possibly explain why Freezer U failed. This seems unlikely because this freezer room had a special lock for which there are only two keys: one held by the tissue processing staff and one kept in a special lock box in the hospital's security department. Additionally, there was a surveillance camera focused on the door to this freezer room and multiple reviews of 4 weeks of tapes showed no evidence of unauthorized entries into this area. At one point in the investigation, it was found that Freezer U started up normally and its compressor system was able to decrease the inner temperature to -78°C!

Conclusions

Although no explanation for the failure of Freezer U was ever identified, the HBTRC has re-doubled its efforts to protect our freezers by evaluating all of our standard operating procedures (SOPs). Additionally, every aspect of our freezer SOPs has been tightened further; each freezer now has three separate alarm systems. The newest one is based on wireless technology and provides sophisticated continuous surveillance monitoring of each freezer; this information can be accessed to identify whether an incipient pattern of dysfunction may be evolving. For all three of these alarms, when the temperature of a freezer unit drops below -70°C, the alarm system will contact 1) the pager and smartphone of the primary on-call person, 2) a backup on-call person who lives very close to the hospital, 3) the person in charge of tissue processing, 4) the associate director, and 5) the director. The latter two individuals are reached via e-mail rather than cell phone or pager. The use of three different forms of technology will increase the likelihood that at least one person in the line-up will be notified of an impending freezer failure and be able to act accordingly.

The HBTRC staff has always understood how critical their stewardship of this precious brain collection is—but even with all the modifications that we have put in place, we will always have a lingering question as to whether another freezer failure, one that defies logic, could ever happen again. The HBTRC is now better positioned to act preemptively if this should occur.

Address correspondence to:

Francine M. Benes, MD, PhD

Harvard Brain Tissue Resource Center

McLean Hospital

115 Mill Street

Belmont, MA 02478

E-mail: fbenes@mclean.harvard.edu

Supported by: NS/NS R24 MH068855

Sanford Health is an integrated health system headquartered in the Dakotas and is the largest rural, not-for-profit health care system in the nation with locations in 126 communities in eight states. Sanford Health includes 35 medical centers, 140 clinic locations, and 1360 physicians, offering expertise in 81 specialties. The Biobank at Sanford Health was established in 2011 as part of an initiative to support genomic research at Sanford Health, and seeks to collect biospecimens from our patient base across the Dakotas.

At the biobank, we identified a need to expand our ultra-low temperature (-80°C) freezer storage as well as a need to build upon our backup systems and therefore decided to purchase multiple ultra-low temperature (ULT) freezer units for both primary storage and backup purposes. In discussions with our vendors, we were informed of a new model of freezer that was not only more economical to run but also had a smaller footprint, while having the same capacity of ones that were currently in use. Given that space and operating cost are always major factors, we elected to utilize these new freezer designs for our backup units.

About 6 months after we placed the new units into service, we began to experience temperature fluctuations followed by a first-stage compressor failure. This occurred on multiple units within weeks of each other. The vendor identified an issue with a manufacturing defect in a specific lot of compressors, and our units were repaired using compressors from a more recent lot where the manufacturing defects had been resolved. However, a few months after these repairs, we again suffered a first-stage compressor failure on one of the repaired units. After this second failure, we were informed that the original compressor design was at fault and were offered a repair using a completely different compressor design that had been tested extensively. We have not had issues with these units since the new compressor design was installed.

Although the compressor failures were surprising given the age of the units, we were well prepared for these unfortunate events. Our freezers are equipped with an independent wireless temperature monitoring system that can provide alerts 24/7 to our staff in the event of temperature fluctuations. Our monitoring system worked well in these instances, alerting us to the increased chamber temperature resulting from the compressor failures. We were also able to analyze the temperature records prior to the failure and identify multiple unexpected temperature fluctuations in the weeks leading up to the failure.

The failure of these units did not affect our primary operations and at no time were our samples compromised due to the multiple backup systems we have in place. Our ULT freezers are not only equipped with the independent monitoring system but also have a liquid CO₂ back-up system in place. The liquid CO₂ system functioned as expected during the compressor failure and maintained temperature in our freezers temporarily. However, the failure of the freezers did leave us in a quandary since we were then in a situation where no reliable backups were available in the event of a failure of one of our primary storage units. We requested, and received, loaner freezers from our local authorized repair company that were used as backup units while we had ours repaired. We also located additional freezers within our institution and negotiated access to them in the event that we required their use. Fortunately, our freezers were repaired quickly and without incident.

We had multiple units of the same model freezer and worked quickly with the vendor to identify which compressor units these freezers contained and to have them replaced proactively rather than wait for a failure to occur. This was despite assurances from the vendor that only some of our units were affected by the defective compressors. We felt that it was important to be proactive about this issue rather than rely on the vendor's long-term field testing. Our suspicions were proved correct when additional units began to fail, including units belonging to other principal investigators within our institution. We believe that the proactive replacement of compressors reduced the number of freezer failures.

The biggest lesson we learned from our experience was to establish and maintain an open line of communication with our vendor. When we first experienced the compressor failures, we maintained daily contact with the sales and technical support groups to obtain the latest information on the problem. When the additional failures occurred, we requested to speak with management who helped to assure us that this situation would be resolved. The management team was able to provide a much more detailed explanation of the issue at hand and provided in-depth information on the proposed resolution. The open dialogue between the vendor management team and me was instrumental in my decision to maintain a business relationship with them and not to switch to a competing product.

We have not experienced any other problems with these freezers since the new compressor design was utilized. However, we continue to monitor our equipment and have developed a number of new operating procedures to manage freezer failures and the potential for needing to take freezers off-line long term. We have instituted weekly maintenance schedules that include a review of temperature logs to spot potential problems, tests of the liquid CO₂ back-up system, and visual inspections of all door seals and compressor filters. In addition, we will perform annual preventative maintenance service visits that include a check of the compressors and refrigeration systems to ensure that they are working optimally.

Address correspondence to:

Chun-Hung Chan, PhD

Director

Sanford Health BioBank

2301 E 60th St-N

Sioux Falls, SD 57108

E-mail: Chun-Hung.Chan@SanfordHealth.org

The National University Health System (NUHS) Tissue Repository (formerly NUH-NUS Tissue Repository) is a disease-based repository with, originally, only one single-specimen collection site at the National University Hospital, Singapore. In mid-year of 2011, the repository was tasked to take over the disease and population-based collection at the Singapore Bio-bank, formerly Singapore Tissue Network, which housed close to one million specimens with collection sites spreading across multiple hospitals and clinics in Singapore. With such a huge national collection under our management, disaster recovery was our main priority when the new facility was set up to house the specimens.

Except for occasional flooding, Singapore is fortunate to be largely sheltered from most natural disasters (such as earthquakes, volcanic disaster, typhoons/hurricanes/cyclones) that affect neighboring countries and the rest of the world. To prevent flood damage, the collection was housed on the upper floors of the building, and the focus turned to risk management of man-made disasters and recovery efforts.

The facility was renovated to be equipped with an adequate power supply; an air-conditioning and fire-sprinkler system; card access (entry and exit); an unauthorized intrusion alarm complemented with closed circuit television; and remote security and response plan. In addition, the unit was equipped with a temperature monitoring system, CO₂ freezer backup and diesel generator backup system, complemented with remote monitoring and a maintenance control program.

The main concern is freezer failure as most of our specimens are stored in -80°C mechanical freezers. Each freezer failure, whether a power failure or mechanical failure, will need to be managed differently. To prepare for a possible blackout, a diesel generator is set up on the rooftop of the building (to minimize flooding risk) providing power during blackouts and annual electrical maintenance of the building (where power will be turn off intermittently). An external contractor is engaged to provide 24-hour generator support when there is a disruption to the regular power supply. A temporary power solution (mobile generator) contract is also in place with a commercial vendor as a backup for the rooftop diesel generator. There are several measures to manage for mechanical failure. All freezers are hooked to a CO₂ tank, have a reliable 24-hour dry ice supply, and, most importantly, empty freezer compartments are available at all times for quick transfer of specimens in the case of extended mechanical failure. There are 4 teams (up to 4 members in each team) on roster duty for providing 24-hour support for any major incident in the facility. At any time, one team is on duty with another as a backup. All members have remote access (via apps on their smartphones) to the CCTV system as well as the monitoring system for forced entry, blackouts, and freezer temperature (both internal and external temperature probes are monitored). E-mails, phone calls, and SMS will be activated when pre-set alert points have been reached.

There are other man-made disasters that cannot be managed by the above measures, such a major fire or collapse of building. To manage these risks, there are plans to split the collection and to house them separately at different geographical sites.

Address correspondence to:

Chon Boon Eng, Ph.D.

Head, NUHS Tissue Repository and Hospital-based Cancer Registry

NUHS Research Office

Yong Loo Lin School of Medicine

National University Health System, Singapore

5 Lower Kent Ridge Road, Singapore 119074

E-mail: Chon_Boon_ENG@NUHS.edu.sg

http://medicine.nus.edu.sg/tissue

Biobanks are moving to innovative dense storage solutions to minimize footprint and energy use and maximize specimen storage and temperature stability. One such system purchased by the Division of Cancer Epidemiology and Genetics (DCEG) and installed in 2011 in the Frederick National Lab (Frederick) Central Repository currently houses over 2.3 million specimens. This 2,500 cu.ft. dense storage -80°C reach-in/-20°C walk-in freezer is comprised of an unrefrigerated anteroom that opens into a central corridor maintained at -20°C that allows technical staff to reach into the -80°C chambers.

Standby emergency pre-cooled backup freezers are cost and space prohibitive for dense storage units of this size and rack configuration. In the event of an emergency, material cannot be easily moved. Therefore, redundant backup systems must be an integral part of system design and emergency planning. To this end, the following measures were taken:

  redundant dual industrial scroll high- and low-stage compressors for both the -80°C space (Side A and B) and dual compressors for the -20°C spaces;

  compressors alternate, each supporting the entire system at required temperature;

  sized so that in the worst case the -20°C system will keep the entire unit at -20°C;

  liquid nitrogen backup plumbed to a bulk nitrogen system for the -80°C chambers, including dual electro-mechanical solenoids and a manual control valve;

  multiple temperature and controller alarm points connected to a 24/7 monitoring system;

  emergency backup power generators; and

  on-site spare replacement parts for critical or consumable items.

In October 2012, the biobank staff experienced and managed a “successful failure” of the system, which occurred when the redundant compressor failed during a planned outage to repair a leak in the primary compressor system. The liquid nitrogen system was activated as designed and maintained temperatures for 24 hours until the redundant compressor could be replaced and tested, and was able to independently cool the unit. Events unfolded as follows.

It began in late September 2012, when staff were alerted to a potential problem when a daily review of temperature over time graphs identified a change in the compressor cycle pattern, indicative of system trouble. A refrigerant leak was discovered on Side A of the system. Control of the system was locked into Side B so that Side A refrigerant could be removed and the leak repaired, and the system tested. Sufficient vacuum was not achieved during testing of Side A on a Friday afternoon, October 5 2012, so the system was charged with nitrogen and left on Side B control.

Staff arrived Monday morning, October 8 2012, to find the Side B compressor making an unusual noise. The compressor then tripped out on thermal overload. Staff tried to troubleshoot the problem and restart the Side B compressor, to no avail. Activity immediately shifted to replacement of the Side B compressor using the on-site spare. Side B refrigerant was moved to the reservoir and additional service technicians were called in to help. At this time, neither Side A or B were cooling the space. Space temperatures rose to -70°C. As programmed, at about 11 am, the backup nitrogen cooling system kicked in to keep the unit at proper temperatures.

Side B repair was under way when it was determined that the stock of refrigerant on site, depleted due to repairs on Side A, was not sufficient to re-charge Side B. There was concern that needed gas could not be delivered for days because the typical large quantities of flammable gas additive could not be transported in service vans due to Department of Transportation regulations. Fortunately, an alternate supplier, with smaller tanks, was identified and the gas received. Alarm points were lowered to -75 °C to alert staff early if there was a problem with the nitrogen-fueled cooling system. On October 9 2012 at 10:15 am, repair and testing of Side B was completed, and Side B was restored to normal function. At that time, Side B took over cooling from the nitrogen system.

Focus returned to the repair of Side A, which was completed over the next few days. Innovative use of dry ice to cool the refrigerant and lower gas pressures enabled technicians to reclaim the system charge within hours rather than days, to allow quick completion of the repairs.

NCI and biobank staff had foreseen and planned for catastrophic failure of the mechanical systems. Liquid nitrogen backup systems and on-site spare parts for critical components led to protection of the specimens and a 24-hour turnaround on replacement of the Side B compressor. If a replacement compressor had not been on site, it could have taken a week or more to get one shipped to the facility. Nontraditional methods to re-charge the system with refrigerant cut days off the repair time.

Lessons learned from this experience include:

  Determined actual rate of nitrogen needed per hour to maintain unit temperatures.

  Identified additional useful spare parts that had not been included in the initial list.

  Refined monitoring checklist to include weekly check of pressures and discharge temperatures for all compressors during normal operation. This allows a proactive response.

  Use of improved fittings on lines to prevent the potential for the kind of leaks that took Side A out of commission.

  Installation of inline isolation valves to improve refrigerant management.

  “Autopsy” performed on the failed compressor indicated a manufacturing defect was the cause.

The importance of advanced planning for emergency situations cannot be emphasized enough: The greater the risk, the more critical the planning. Risks need to be assessed and either avoided, transferred, mitigated, or accepted. Emergency situations can be controlled only if resources, including experienced staff, spare parts, materials and supplies, tools and equipment, are on site and able to be used. Identification of failure points up- and downstream of the system is the first step. Implementation of redundant and backup systems to address these points will provide the greatest chance of successful management of a failure situation. Unforeseen situations occur even with the best planning, so failure plans should be updated following an event to incorporate lessons learned.

Address correspondence to:

Kathleen H. Groover, PhD, CIH

Project Director

NCI Frederick - Central Repository Services

Fisher BioServices

4600 Wedgewood Boulevard, Suite H

Frederick, Maryland 21703

E-mail: kathleen.groover@thermofisher.com

Southern Health is one of the founding members of the Victorian Cancer Biobank in Australia. At this site, biospecimens collected from public hospitals within our region are processed and stored.

During each calendar year, Southern Health Tissue Bank collects and stores an average of 19000 samples–of which at least 85% require storage within a cryopreservation facility. We have installed two freezers for -80°C storage and liquid nitrogen storage at -190°C. These are based at our Clayton facility and are used as permanent storage facilities. In addition there is a -80°C temporary storage facility at the Moorabbin campus which is approximately 7 km from the main Clayton campus.

During the past 18 months, we have had two freezers break down. Both freezers were of 700L capacity and were connected to the hospital building maintenance system. This system is established to record and alert the security department in case of freezer/liquid nitrogen storage failure. The alarm is triggered if the freezer temperature rises above -70°C. The security department at Southern Health has a contact list of all tissue bank staff members with their names, pager and mobile numbers. The list was designed as a cascade so that security will aim to contact the personnel that lives closest to the facility before moving to the next contact.

The first event of the failure happened in the early morning. The security failed to alert the tissue bank staff that the freezer was losing its temperature level. The temperature rise was noticed by a staff member when checking the freezer temperature panel and noticed the discrepancy between the additional monitor on the outside of the freezer and the actual freezer temperature panel. The discrepancy in reading was over 20°C and seemed quite abnormal. It was later discovered that during that morning the hospital had an intermittent power failure which most likely triggered the freezer failure; although it was plugged into the generator supported power point.

The second event was related to a construction project at our facility to improve liquid nitrogen storage and to provide a ventilated environment for the freezers. Unfortunately, the workers closed the door before the air conditioning was turned on, as a result of which the room overheated. The freezer compressor could not sustain the additional load and hence stopped functioning. In this case, the building maintenance alarm was triggered. The security staff then alerted us of the freezer failure.

Fortunately, in both cases, action was taken to avert sample disintegration. The temperature had risen to -40°C in the first case and to -60°C in the second case. The integrity of samples was jeopardized. We had to move our samples to the backup storage within the building. Since we could not accommodate all the samples at one site, some samples were placed on dry ice and moved to another site.

To empty and cart samples from a 700L freezer, which can store up to 44000 samples, takes about one hour and a half if the backup site can be reached within 5 minutes. It takes more time if samples need to be placed on dry ice and carted to another facility altogether. Removal of samples requires a minimum of two staff members.

We faced major issues not only to locate where to store the samples in order to preserve their integrity but also to organize enough staff to help with the sample move. The timing was also important as freezers do lose their temperature fairly quickly, particularly when one has to open the door in order to retrieve samples.

In our case, we did not have a backup facility dedicated to the tissue bank. We had, however, negotiated with the diagnostic pathology department and the clinical trials department so space could be allocated to us should we need it. In both cases, we did not have enough dry ice stored at the tissue bank and we had to rely on our biochemistry department for dry ice.

Even when you believe that every reasonable step has been taken to secure your repository; it may not be the case.

In the first event described, the hospital security failed to alert us of the failure, and the event was discovered almost by chance. Both freezers are plugged into the generator-supported power points in case of freezer failures. And in the second event power was not the issue. Following these Events we had to try to improve our security levels to protect the samples from storage facility failure.

In order to mitigate such events we had to go through the preventive methods again. We had detailed discussions about these issues with our peers who are in a similar situation. Suggestions from them consist of having CO₂ cylinders connected to each freezer, which allows another 72 hours for removal of samples; to invest in a “boost power box” that can provide power to the freezer for 30 minutes after a failure, preventing power-induced freezer failures. When considering temperature monitoring systems through building maintenance systems keep in mind that any building maintenance work could affect its reliability for that period of time.

It is reasonable to conclude that the majority of man-made disasters are mechanically induced and require a permanent monitoring system. One monitoring system, however, is insufficient. When managing a facility of this kind it is important to be informed about any building projects current and prospective, to request reports on power failures from the engineering department, to do thorough and systematic visual checks on all temperature monitoring panels, to have an additional temperature tracking device on each freezer that allows for a printed temperature record. This additional temperature tracking can be set up to remotely alert your pager in case of any malfunction. It is also important to ensure that there is a backup freezer space in near proximity of the existing facility, which will reduce sample removal time and enable preservation of valuable samples.

Address correspondence to:

Zdenka Prodanovic

Tissue Bank Manager

Southern Health

A member of Victorian Cancer Biobank MMC Clayton

246 Clayton Road

Clayton, Victoria 3168 Australia

MMC Clayton

246 Clayton Road

Clayton, Victoria 3168 Australia

E-mail: zdenka.prodanovic@southern.health.org.au

The responsibility to safeguard precious samples is a key component of running a successful biobanking operation. No matter how diligently the plan for disaster recovery has been considered and prepared for implementation, there is always the possibility of an issue arising that was completely unexpected.

Since the inception of the CTSI Biorepository at the University of Florida, our goal was to gradually build the infrastructure and processes needed to safeguard the samples. Due to economic constraints, this can sometimes be challenging. However, we propose that there are minimal requirements that are necessary from the start of the biobank with low economic impact, which can be expanded to more sophisticated infrastructure as budgets allow. The following is what we would consider the minimum necessary equipment that must be in place:

  CO₂ or LN₂ back-up on every storage unit as applicable;

  monitoring systems;

  storage temperature settings;

  room temperature/humidity settings;

  backup emergency power with load testing performed on a routine basis;

  at least one completely free storage unit of each type or 10% backup capacity, whichever is greater;

  splitting collections into different storage units and duplicating collections at off-site storage locations whenever possible.

The minimum processes that must be in place are as follows:

  emergency response plan for internal and external disasters with regular training of staff members;

  establishing relationships internally and externally (possibly within a nearby state) for temporary storage (minimum of 3 months) of samples.

When budgets allow, expanding on the minimum necessary infrastructure is crucial and may be done as described in more detail here. Regarding CO₂ and LN₂ backup, it is important to set up a contract to have access to replacement tanks on a 24/7 basis. It is anticipated that this will cost more than the standard contract to have the tanks replaced during regular business hours. For monitoring systems, there are two relatively inexpensive ways to proceed. The first is to use an external NIST traceable thermometer (important to track the expiration date) in each system and to check the temperatures at least twice daily. The second is to use a remote monitoring system that utilizes the temperature probe of the storage unit to monitor the temperature. If this probe fails and continues to read a temperature within acceptable limits, it is likely that the failure would not be discovered until it is too late. The more expensive and technologically advanced systems (allowing for remote web access, etc.) use temperature probes external to the system, providing a redundancy in the temperature monitoring. In addition, the temperature should be recorded twice a day for each storage unit including on weekends and holidays. Due to the cost of having staff work off-hours with potential overtime, this can sometimes be cost prohibitive. However, with the remote access systems, the cost of this can be offset by recording that temperatures were remotely verified. Regarding backup emergency power, it is important to determine your biobank's priority with regard to the provision of power within your organization, as well as to determine the length of time it is estimated that this backup emergency power will last in the event of a serious disaster with extended loss of power.

Regarding the disaster plan itself, it is important to develop this with consideration for all possible disaster-related events. We recommend having processes in place for both internal (occur within the biobank itself due to a freezer failure, etc.) and external (man-made or natural) disasters. The disaster plan should clearly outline the steps to take at all stages of the event, allow for problem solving in the event that the steps cannot be followed, require documentation of all samples impacted by the event (including individual barcode number whenever feasible), and filing of deviation reports following the event with updates once the recovery has taken place (i.e., the samples have been returned to their proper locations within the facility). In addition, some general considerations for both internal and external disasters include:

  regularly scheduled training on the disaster plan, trying to mimic the events and reactions that would be required;

  establishing relationships with internal (i.e., other labs in your department or outside of your department if in an academic setting) and external groups (i.e., outside academic institutions within and/or outside your state if in the United States) to use available storage space;

  establishing contracts with vendors who can help transport samples to and from a new location samples at the required temperatures in advance of a disaster;

  creating a priority list of samples to be moved; and

  establishing a supply of a cash to be available in the event of power loss, to purchase items such as dry ice, etc. Such an event may prohibit the use of credit cards or ATMs.

Regarding the priority list for samples mentioned above, it is likely that in an external or expansive internal disaster such as entire building loss of power for an extended period of time, it may be difficult to attend to all of the samples that have been impacted. This priority list should be posted in the laboratory along with a list of contacts in emergency situations that include service providers, vendors, owner(s) of the collections, etc.

Background for CTSI Biorepository for the University of Florida

Fortunately, since the CTSI Biorepository's inception 4 years ago, there have been few disasters to deal with. Of three events, all were internal and of little consequence. The first was complete loss of a single freezer that contained no samples and was due to a manufacturing default. The second was a brief loss of power to the entire building for no determined reason, but the event provided an additional opportunity to ensure that our backup power was working. The third was planned loss of chilled water used to cool the HVAC system on two occasions (one brief and one overnight due to construction), and portable cooling units were placed around the freezers by the University's physical plant division to maintain room temperature. The true test of how well we have prepared for a disaster will come, most likely in the form of a large-scale power loss due to a hurricane, with potential to impact not only the loss of power to freezers but the loss of air conditioning that keeps our freezers from overheating due to the humidity and heat in Florida. One of the most important considerations for the disaster plan itself is to design the plan based on the most likely disaster for your geographical location. In Florida, storms with lots of rain and wind that could cause downed power lines and lightning that could take out transformers even when these storms are not on the scale of a hurricane are probably the highest disaster risk that might prevent safeguarding of the irreplaceable samples.

What Is in the CTSI Biorepository Disaster Plan?

The CTSI Biorepository currently houses eight -80^o freezers all of which are connected to a CO₂ backup and two remote monitoring systems (the basic and more sophisticated monitoring systems) to allow dual redundancy in temperature monitoring. Each freezer also has an external NIST traceable thermometer. The remote monitoring systems are set up to call out to at least three individuals and will cycle through those numbers until it obtains a response acknowledging the message was received. One of the monitoring systems is verified daily to ensure that they are operating correctly and call out appropriately. All of the freezer systems are on emergency backup power in the event that there is power loss to the building where they are located. The CTSI Biorepository also has an emergency plan SOP and associated forms. The forms include 1) emergency contact list posted in a visible location on all freezers and in a number of outside entry points 2) emergency response checklist and 3) samples affected by freezer malfunction. A deviation report will be completed after each emergency. The CTSI Biorepository conducts group training on the emergency response plan for all staff every year. We are currently working on expanding the emergency plan SOP to better prepare for external disasters. In terms of prioritization of samples, the CTSI Biorepository stores both investigator samples as well as samples collected as part of an internal library. In an emergency situation, the investigator samples would have priority over the internal library collection. Although these decisions are often difficult to make, they must be considered when time is of the essence in protecting such a valuable resource.

Melissa Rawley-Payne MA¹; Rosie Kizza MHA, RN²;

Chris Hiam, BS²; Amer Abouhamze, MHA²

¹MRP Biobank Consulting, LLC

²University of Florida, CTSI Biorepository