One Resignation Away: What Leaves the Lab With a Person
The SOP says what to do. The person who left knew what actually happens: the second warm-up, the client's report format, the Tuesday calibration nobody wrote down. The next assessor asks for those.
In brief
What leaves a lab when an experienced person leaves? The method stays in the SOP. What leaves is the exceptions: the instrument that needs a second warm-up, a client's report format, and the checks one person did on a schedule from memory. Those become the records an assessor finds missing, and an afternoon's audit finds them first.
Key takeaways
- 76% of MLO's 2026 respondents report high or moderate turnover, and 52% are 56 or older, so more departures are coming.
- The SOP holds the method. The exceptions live with the person or in a spreadsheet nobody checks, and they break first.
- CAP's top deficiency drivers include missing records for scheduled activities, often the work one person did from memory.
- Start the audit in one afternoon. List every scheduled activity, name who does it, then name who else can do it.
- Exceptions belong on the record beside the instrument, method and client, where the next person will look for them.
The microbiologist’s last day is a Friday. The lab buys a cake, and she leaves a two-page handover memo on your desk. The next Tuesday, the incubator calibration she has done every week for nine years doesn’t happen, and nobody notices. It was never on the QA calendar, because she had always just done it.
In March a client calls about an EDD. It came in the wrong format, and the person who knew how that client wanted it has been gone five months.
You already know about the lab staffing shortage. In MLO’s 2026 salary survey, 76% of respondents reported high or moderate turnover, and 52% of them were 56 or older, so more of those Fridays are coming. This article skips the hiring advice. It covers what leaves with a person, why the SOP misses it, and how to find it before the next resignation, whether your lab runs on binders or a LIMS.
What actually leaves when an experienced person leaves?#
The method stays in the SOP, and your replacement can open the plate reader procedure and run it on day one. The SOP leaves out the rest. The plate reader needs a second warm-up on cold mornings before the blank passes. One client wants the lab’s sample ID in the first column of the EDD. The Tuesday calibration lives in one person’s head. These exceptions are what walk out the door when she leaves.
Some exceptions are written down, but in places nobody looks. When we move a lab onto a new system, the costly ones usually turn up in a spreadsheet or inside a report calculation.
In a typical food lab, a spreadsheet ran the spec checks for the two biggest customers and trended “<10 CFU/g” as 10, so one customer’s charts showed counts that were never there. In a typical water utility, the monthly report averaged non-detects as zero, but one plant’s permit called for a different rule.
A lab can be “one resignation away from total failure,” Stacey Metzler of Phoenix Children’s told Lab Manager’s 2026 Leadership Summit. Lab Manager’s report of her talk says that when calibrations “exist only in a single technician’s memory,” “unexpected staff turnover creates an immediate bottleneck.”
Your new hire can learn the method from the SOP, but nobody is left to teach the exceptions.
Why do SOPs not capture it?#
Nobody raises a document change for something they fix with their hands at 7 a.m. The microbiologist found the problem on a cold morning in 2019 and solved it the same morning. The SOP said 30 minutes of warm-up, and on cold mornings 30 minutes wasn’t enough. She knew that, but nobody ever added it to the SOP.
Most exceptions start that way. Someone hits a problem on an ordinary morning, fixes it before the run and gets on with the day. The fix works, so it turns into a habit, and the habit never reaches the manual.
An SOP is written for the normal day, and changing it takes a formal step. CLIA requires every procedure change to be “approved, signed, and dated by the current laboratory director before use.” That rule is right, and it’s also why the second warm-up never made it into the manual.
Tywauna Wilson, writing in Lab Manager this year, recommends “Annotated SOP reviews that flag every step where real practice differs from what is written.” Do it at the bench, with the person, one method at a time. We would rather see a messy instrument note from 2019 than a clean SOP that never mentions the problem. That note holds what the person at the bench knows.
What do assessors find missing after a departure?#
Assessors find the records for activities done on a schedule, because those are the records that stop. In November 2023, CAP told CLIAC what sits behind its top ten accreditation deficiencies. The first common reason it listed is “Missing records for activities to be performed at defined frequencies,” from daily checks to annual ones.
The same presentation names “Workforce Shortages and Inexperienced Staff (Staff turnover)” as a root cause. CAP doesn’t connect the two, but the microbiologist’s lab shows how one leads to the other.
The Tuesday calibration log stops the week she leaves. The calibration was never on the QA calendar, so nobody goes looking for it, and by the time the assessor opens the binder, eleven weeks have no entry. The assessor can’t tell “done and not recorded” from “not done” and cites the gap either way. Maintenance and function checks, instrument record review, temperature checks and competency assessment were all in CAP’s top ten, and each one recurs on a calendar.
How do you find the single points of failure before the resignation letter?#
Start by counting how many people can do each task. Wilson calls a task only one person can do a “single point of failure,” and she starts cross-training with competencies that “currently have fewer than two qualified practitioners.”
You can do a first pass this week. Spend one afternoon with the people nearest retirement or whose names come up most. Start with the QA schedule, then ask each of them what they do that isn’t on it. The microbiologist’s lab would have found four single names in an hour: the Tuesday calibration, the warm-up, the client’s EDD and the PT portal login.
Expect the list to run long. In a typical migration, a lab’s list of five export formats and 80 reports turns out to be nine formats and 124 reports.
- List every scheduled activity, from calibrations and maintenance to QC review, PT and competency.
- Add every client with a preference, such as an EDD format.
- Add every instrument with a quirk the manual leaves out.
- Add every export, integration or login that runs under one person’s account.
- Add every spreadsheet or report that does a calculation the system doesn’t, such as a spec check or a non-detect rule.
- Add every judgment call only one person makes, and every CLIA role one person holds.
- Write a name, a second name and the next due date beside each line. Count the lines with one name, and do the daily ones first.
An export set up under one analyst’s login is an integration that runs as somebody, and it stops the day that account is disabled.
The lines with one name are your cross-training plan, starting with the ones due soonest.
Where should the exceptions live?#
Put each exception on the record it applies to. The warm-up goes on the plate reader’s record, and the EDD format goes on the client’s record. A spec check in someone’s spreadsheet belongs in the system that reports the result, where everyone can see it. A last-week memo can’t hold all of that, because its writer doesn’t know which habit you’ll need in March.
| What the person knew | Where it lives | The rule |
|---|---|---|
| Maintenance and calibration | The instrument’s log | 42 CFR 493.1254; 21 CFR 211.182 |
| The second warm-up | A note on that log | Practice |
| A method change and why | The method’s version history | 42 CFR 493.1105(a)(2); 21 CFR 211.194(b) |
| A client’s report format | The client’s record | Practice; 42 CFR 493.1291(e) for updates |
| Who did, checked or may do it | Each record; the personnel file | ISO/IEC 17025 7.5, 6.2; 42 CFR 493.1407(e)(14) |
Everything else comes from time at the bench with the person who knows it. Ask that person to own the list and pick who learns each line, so it reads as a request to teach instead of a hint to leave. Have the backup do the task on the day it’s due while the senior person watches, and write what they learn on the record.
CLIA’s technical supervisor already has to observe instrument maintenance and function checks to evaluate competency. Whether the record is a bench logbook or a laboratory management system, it should be the first place the next person looks.
Where LabLynx fits#
<!– CONFIRM Claims Laurie, John or Brandon must confirm before this goes past staging (revised after the reader panel, 2026-10-19): 1. Section 6: "When we bring a lab onto LabLynx LIMS, the longest conversation is about the exceptions nobody wrote down." (Laurie: "true to what I see"; John to sign.) 2. Section 6: "In a typical migration, those are the client EDDs missing from the lab's own list, the spec checks running in a spreadsheet and the report calculation that treats non-detects as zero." (Laurie: her composite stories 1, 3 and 4, labeled typical.) 3. Section 6: "We move them into the system, where the next person can see them and every change shows who made it and when." (Laurie: stories 2 and 3 moved spreadsheet and report logic into the LIMS; John: the change history is in the live how-does-a-lims-work article, the rest is not in any live LabLynx text.) 4. Section 6: "If you'd like a second opinion, bring the list to a call with us. We'll tell you which lines a system can hold and which need a second person, and the list stays with you." (Brandon: that sales takes that call and offers it on these terms; it is the only call to action.) 5. Section 1: "When we move a lab onto a new system, the costly ones usually turn up in a spreadsheet or inside a report calculation." (Laurie: her field note on this draft.) 6. Section 1: "In a typical food lab, a spreadsheet ran the spec checks for the two biggest customers and trended "
Frequently Asked Questions #
What is tribal knowledge in a laboratory?
Tribal knowledge is the know-how a lab runs on that nobody has written down. It is the instrument that needs a second warm-up, the client who wants the EDD a particular way, and the calibration one person does from memory every Tuesday. SOPs cover the normal case and change through a formal approval, so these exceptions stay in people's heads until someone records them.
How long does it take to replace an experienced lab analyst?
No current survey publishes a figure, so plan on the vacancy instead of a date. ASCP's 2024 Vacancy Survey, with 1,027 responses covering 18,626 employees, found vacancy rates lower than in 2022 but above pre-pandemic levels, anatomic pathology highest at 28.5%, and retirement rising in 10 of the 17 departments surveyed. In MLO's 2026 survey, 79% of respondents reported difficulty hiring.
What should a handover document include when a lab scientist leaves?
Start with every scheduled activity the person did, such as calibrations, maintenance, QC review and PT submission, with its next due date. Then list the exceptions: instrument quirks, client reporting preferences and sample naming habits. Then every login, integration, spreadsheet and report calculation that runs through that person. Put each item on the record it belongs to and keep the memo as the index.
What is a single point of failure in a lab?
A single point of failure is one person, instrument or file the lab cannot run without. In staffing it is a scheduled activity or critical competency only one person can perform, such as a calibration that exists in one technician's memory or the only login on the PT portal. You find them by listing every recurring activity and counting the lines with one name.
Does ISO/IEC 17025 require records of who does what?
Accreditation bodies that assess to ISO/IEC 17025 expect it. As they reproduce clause 7.5.1, a technical record carries the date and the identity of the person responsible for each activity and for checking the results, recorded when the work is done. Clause 6.2 covers personnel competence, authorization and monitoring records, and clause 8.4 covers control of records.
How do you cross-train lab staff without stopping the work?
Start with the activities that have a date, such as the next calibration or the PT submission. The second person does the task on the day it falls due while the first person watches, so no bench time is added. One MLO respondent named the constraint: "I cannot take time away from the lab to attend." Training on the day the work happens answers that.
Sources and references
- ISO/IEC 17025:2017 General requirements for the competence of testing and calibration laboratories, clause 6.2 Personnel International Organization for Standardization
- 21 CFR 211.25 Personnel qualifications Electronic Code of Federal Regulations
- The American Society for Clinical Pathology 2024 Vacancy Survey of medical laboratories in the United States American Journal of Clinical Pathology
