Measuring an HTM Service: Which Metrics Matter and How to Compare Them

Every healthcare technology management service reports something, and most report too much. The difficulty is rarely a shortage of numbers. It is that two departments can quote the same metric, mean different things and both be correct within their own definitions.

Definitions Decide the Result

A metric is a sentence with a measurement attached, and the sentence is the part that matters. Device uptime measured against the working day produces a different figure from uptime measured against continuous clinical demand, and neither is wrong until someone compares them.

Before any number is published, the service should be able to state what is counted, what is excluded, where the data comes from and who owns it. That statement is more valuable than a target, because a target without a definition cannot be met or missed in any meaningful way.

The Metrics Worth Carrying

A small set, reported consistently, tells a better story than a dashboard of dozens. The four families below cover responsiveness, reliability, discipline and cost, which between them answer most of the questions a hospital executive will ask.

The pattern in the third column is worth more attention than the numbers in the first. Almost every metric can be improved by changing where the clock stops, and a service that does not audit its own definitions will eventually be audited by someone less sympathetic.

Uptime and Availability

Uptime sounds unambiguous until it is applied to equipment that is shared, pooled or held in reserve. A infusion pump in a store room is available to the organisation and unavailable to the ward that wants it, and both statements can be true at the same time.

Availability is therefore better expressed against clinical need than against the calendar. A device that is out of service during a scheduled list is a different problem from one that is out of service overnight in an empty department, even when the elapsed time is identical.

Some services separate device availability from clinical availability for exactly this reason. The first measures the fleet, the second measures whether a clinician could start the case, and the gap between them is usually a logistics problem rather than an engineering one.

Response, Repair and Resolution

Response time measures the interval between a request and an acknowledgement. Resolution time measures the interval between a request and a working device. They answer different questions and should never be reported under one label.

A fast response with a slow resolution may indicate a parts problem or a skill gap. A slow response with a fast resolution may indicate triage that is working well but understaffed at the front end.

Add a third measure for repeat visits to the same device, because a fault that returns is more costly than a fault that was fixed once and stayed fixed. Repeat repair rates expose training needs and diagnostic shortcuts that averages hide.

Planned Work as a Leading Indicator

Completion of planned maintenance is the metric most closely tied to future performance, since it describes work that prevents failures rather than work that responds to them. It is also the easiest to game, because a work order can be closed without the task being done.

Useful reporting therefore pairs completion with evidence: who performed the task, what was found and whether any deviation was raised. A completion figure supported by signed records and a small number of findings is credible. A perfect completion figure with no findings at all suggests the records are being produced rather than the work.

• Completion by device family, so a neglected group does not hide inside a fleet average.
• Overdue work aged by how long it has been overdue, not simply counted.
• Findings raised per completed task, as a rough indicator that inspection is real.
• Backlog in hours, which converts a queue into a staffing conversation.

Benchmarking Without Fooling Yourself

External comparison is useful when the comparison is like for like and dangerous when it is not. Services differ in case mix, building age, whether they cover imaging, whether they manage information technology and how much work is contracted to manufacturers.

The conditions for a fair comparison are unglamorous. The same definitions, the same inclusion rules, a similar scope of responsibility and a similar split between in-house and contracted work. Without those, a difference in the metric may be a difference in the paperwork.

Comparison is most valuable when it is used to ask a question rather than to award a verdict. A service that appears slower than its peers should be able to explain why, and the explanation is often a legitimate difference in scope that no benchmark captures.

Reporting That Changes Something

Metrics earn their place by prompting a decision. If a report never changes a staffing plan, a contract or a maintenance interval, it is decoration, and decoration consumes the time of the people who produce it.

A workable rhythm is a small monthly view for the service itself, a quarterly view for operational management and an annual review that connects performance to the capital plan. Each audience needs a different level of detail, and none of them needs every metric.

Measure, Explain, Adjust

The strongest HTM services treat measurement as part of the work rather than as a report about it. They define terms narrowly, publish the definitions alongside the figures and are willing to say when a number is not comparable.

That habit connects directly to the rest of the discipline, from the preventive maintenance schedule that generates much of the data to the workforce planning that the metrics are meant to inform.