HA 640 Unit 8 operations improvement recommendation example

From 130 to 195 Tubes an Hour for $47,756 a Year: An Operations Improvement Recommendation for Morning Chemistry Results, With Paired Measures and a Stated Test of Failure

Student Name

Health Care Administration, Herzing University

HA 640: Healthcare Operations Assessment and Decision-Making

Instructor Name

March 5, 2026

From 130 to 195 Tubes an Hour for $47,756 a Year: An Operations Improvement Recommendation for Morning Chemistry Results, With Paired Measures and a Stated Test of Failure

Recommendation

Put a third specimen processor on the bench every morning from 4:30 to 8:30 starting in week 8, and implement bedside barcode labeling with a result-status message over 27 weeks, raising morning processing capacity from about 130 to 195 tubes an hour and cutting the median wait before accessioning for peak tubes from 31 minutes to about 4, at a cost of $135,756 in the first year and $47,756 a year after that.

The Numbers Behind It

Baseline. Over 28 days in November 2025, peak-window routine chemistry tubes needed a median 88 minutes to reach a verified result, with a 90th percentile of 141 minutes; tubes arriving at other hours took a median of 41 minutes. A commonly suggested goal is 90% of routine results within 60 minutes (Hawkins, 2007).

Constraint. The processing bench, whose two morning processors clear roughly 130 tubes hourly, is the only stage whose capacity falls below the morning arrival rate of about 152 an hour. The centrifuges and analyzers run at under half their capacity during the peak.

What the process can produce while the constraint holds. A ceiling of roughly 260 processed in the 5-to-7 window, while some 305 come in, leaving a backlog of about 45 tubes at 7:00 a.m. that clears after 8:00. The forecast projects the peak arrival rate at 139 to 160 tubes an hour throughout 2026, so the constraint binds all year and tightens slowly as volume grows.

Options Considered and Why They Fell Short

A third chemistry analyzer. Rejected: it adds capacity at a stage already idle for more than half the peak, and would not change the morning result time at all. Strengthening a non-constraint does not raise the output of the system (Goldratt & Cox, 2004).

Doing nothing. Rejected: it scored lowest in the decision matrix, leaves physicians without results for about a third of patients at rounds, and the backlog grows in winter.

Barcode labeling alone. Not sufficient: its capacity gain, to roughly 141 an hour, leaves the bench short of the peak. It is part of the recommendation for its quality benefit, not as the capacity fix.

Pre-analytic automation. Deferred: it would solve the capacity problem permanently but costs about $102,000 a year when capital is spread over its life, takes eight to ten months, and ranked below the third processor on the agreed weights. It becomes the right choice if volume grows past what three processors can handle, which the forecast does not project within two years.

Splitting morning draws. Deferred, not rejected: it would cost less and would let stable patients sleep through the early draw, but it depends on medical staff agreement about which patients qualify. The laboratory director will bring it to the medical executive committee in April; if approved, it could replace the third processor's early hours in the second year.

Implementation

Weeks 0 to 2: budget approval, laboratory director. Weeks 2 to 8: recruit and train the third processor, talent acquisition partner and laboratory supervisor; the processor begins in week 8. Weeks 2 to 11: select hardware and build the bedside workflow, information services. Weeks 11 to 21: integrated test, two-unit pilot, training and go-live on all 22 units, information services with nursing. Weeks 2 to 6: the status message, built and switched on by a second analyst. Weeks 21 to 27: measurement, laboratory quality coordinator. The path with no slack runs through the record build, the pilot and training, so the information services analyst's time is reserved in writing when the budget is approved.

Cost and Return

First-year cost is $135,756: $41,756 for the processor shift, $88,000 for printers and scanners, and $6,000 for supplies and support. From the second year the cost is $47,756 a year. The return is not mainly in laboratory dollars. It appears in earlier morning decisions: potassium and anticoagulant adjustments made at rounds rather than after, and discharges that no longer wait for a morning panel. The hospital's case management office estimates that about 12 discharges a week are held for morning laboratory results; if half of those left two hours earlier, the beds would be available sooner for admissions waiting in the emergency department. Those benefits will be tracked but are not counted in the decision, which rests on turnaround alone.

Measures in Pairs

Each improvement measure is paired with a measure of what the change might damage.

What Would Count as Failure, and When

Twelve weeks after the third processor starts, the laboratory director will compare the morning measures with the baseline. The change will be judged to have failed if the median receipt-to-verified time for peak tubes is still above 70 minutes, or the 90th percentile above 110 minutes, on the four busiest weeks of that period. Failure would mean the constraint had moved or the estimates were wrong; the response would be a new bottleneck assessment within a month and, if processing is still the constraint, a decision on automation. For labeling, failure is an exception rate still above 1.3% six months after go-live, which would trigger a review of compliance by unit before any further spending. The evidence that barcoding reduces identification errors is strong (Snyder et al., 2012), so a failure there would point to implementation rather than to the idea.

Success, by contrast, would be visible to physicians in weeks: results for most morning draws posted before rounds begin. The laboratory will report the first month's results to the medical staff directly, because the people who raised the problem should be the first to see whether it has been solved. The laboratory and figures are composite.

References

Goldratt, E. M., & Cox, J. (2004). The goal: A process of ongoing improvement (3rd rev. ed.). North River Press.

Hawkins, R. C. (2007). Laboratory turnaround time. The Clinical Biochemist Reviews, 28(4), 179-194.

Snyder, S. R., Favoretto, A. M., Derzon, J. H., Christenson, R. H., Kahn, S. E., Shaw, C. S., Baetz, R. A., Mass, D., Fantz, C. R., Raab, S. S., Tanasijevic, M. J., & Liebow, E. B. (2012). Effectiveness of barcoding for reducing patient specimen and laboratory testing identification errors: A Laboratory Medicine Best Practices systematic review and meta-analysis. Clinical Biochemistry, 45(13-14), 988-998. https://doi.org/10.1016/j.clinbiochem.2012.06.019