Time-Driven Activity-Based Costing

Time-Driven Activity-Based Costing for Care Pathways

Time-driven activity-based costing (TDABC) works out what a piece of health care costs by following a patient through each step of care, recording how many minutes of each nurse, doctor, room or machine that step takes, and pricing each minute at the cost of keeping that resource available for work. Robert Kaplan and Steven Anderson set it out in a 2004 Harvard Business Review article as a simpler form of activity-based costing, and Kaplan and Michael Porter set out a seven-step version for health care in 2011. The method now supplies patient-level costs for micro-costing studies and economic evaluations, and its separate count of unused capacity bears directly on what a change to a care pathway would actually save.

Two estimates per resource in place of time surveys

Conventional activity-based costing asks staff what share of their time goes to each activity and uses those shares to move resource costs into activity pools. Kaplan and Anderson report that employees asked this question invariably give percentages that add up to 100, so time spent idle disappears into the cost of the work that was done.

TDABC drops the survey. For each resource group it needs only the cost per time unit of supplying capacity and the time each unit of work consumes, so clinicians are asked how many minutes one visit takes. The shared logic of tracing costs is covered under cost allocation.

Capacity cost rates built on practical capacity

The capacity cost rate turns a budget line into a price per minute, computed separately for each resource group so that a minute of a consultant and a minute of a healthcare assistant carry different prices.

$$\text{CCR}_r = \frac{C_r}{Q_r}$$

where $\text{CCR}_r$ is the capacity cost rate of resource group $r$ in GBP per minute, $C_r$ is the total cost of supplying that resource for the period (salary and employer on-costs for staff, or the annual cost of space and equipment), and $Q_r$ is its practical capacity in minutes over the same period.

Practical capacity is the time a resource is actually available for productive work. Kaplan and Anderson suggest, as a rule of thumb, 80% to 85% of theoretical capacity, with about 80% for people to allow for breaks, arrival and departure, communication and training, and a 15% allowance for machines to cover maintenance, repair and scheduling fluctuations. They argue that the aim is to be approximately right, within 5% to 10%, rather than precise.

Some health care studies build practical capacity from contracted hours. Crocker-Buque and colleagues, costing vaccination appointments in nine English general practices, divided each staff group's salary and on-costs by available working minutes, assuming 45 working weeks of 2,250 minutes (37.5 hours) for full-time clinical time. Shakya and colleagues' 2025 systematic review found eight studies that calculated practical capacity without allowing for breaks, holidays and training days. Overstating capacity lowers the rate and understates every activity cost that uses the resource.

Minutes per activity and time equations

The second estimate is time. Kaplan and Anderson stress that the question is how long one unit of an activity takes, not what percentage of the day goes on it, and that the answer can come from interviews or direct observation. In health care the times are attached to a process map of the clinical pathway, and Crocker-Buque and colleagues had clinical and administrative staff keep activity logs for two weeks.

Where duration depends on the case, a single average time hides the variation that drives cost. Kaplan and Anderson's time equations show how characteristics of the case make processing times vary.

$$t_{ar} = \beta_0 + \sum_{k=1}^{K} \beta_k X_k$$

where $t_{ar}$ is the minutes of resource $r$ used by one occurrence of activity $a$, $\beta_0$ is the base time for the simplest case, $X_k$ equals 1 when case feature $k$ is present (for example a first visit or the need for an interpreter) and 0 otherwise, $\beta_k$ is the minutes that feature adds (negative if it saves time), and $K$ is the number of features.

The cost of a patient's episode sums the time-based costs over every activity and resource on the pathway and adds the items not consumed by the minute.

$$C_p = \sum_{a} \sum_{r} \text{CCR}_r \, t_{ar,p} + M_p$$

where $C_p$ is the cost of the care episode for patient $p$ in GBP, $t_{ar,p}$ is the minutes of resource $r$ used in activity $a$ for that patient, $\text{CCR}_r$ is the capacity cost rate defined above, and $M_p$ is the cost of consumables, drugs and other items charged per unit used.

The seven-step and eight-step frameworks in health care

Kaplan and Porter proposed patients and their conditions over a full cycle of care, not organisational units, as the unit of cost analysis. Their seven-step framework, as tabulated by Shakya and colleagues, begins with the medical condition and its care delivery value chain and ends with capacity cost rates and the total cost of patient care. In 2019 da Silva Etges and colleagues proposed an eight-step version for micro-costing studies, which opens with the study question or technology to be assessed and ends with a step of cost data analysis. Of the 32 studies in the Shakya review, which left out surgical costing studies, seven were full economic evaluations and 27 followed the seven-step framework.

Unused capacity and what a pathway change saves

Because rates are set on practical capacity, TDABC assigns to patients only the capacity they use, and the remainder appears as a separate line for unused capacity. In Kaplan and Anderson's customer service illustration only 83% of practical capacity was used for productive work, so only about 83% of the department's expenses were assigned to its outputs.

This split is the method's most useful feature for economic evaluation. A TDABC unit cost is a full cost per minute used, so it measures an average cost and can sit well above the marginal cost of one more patient. A pathway that removes minutes frees capacity, and the incremental cost saving becomes cash only if the freed minutes lead to fewer staff hours or sessions, and becomes extra output only if more patients are treated. Otherwise the saving reappears as unused capacity, valued at the opportunity cost of what the freed time could do. Crocker-Buque and colleagues draw the same line for premises, noting that practice facility costs are fixed and unchanged by the vaccination programme.

Worked example costing a nurse-led clinic (illustrative)

All figures in this example are illustrative. A nurse-led outpatient clinic uses a nurse, a healthcare assistant for observations and a consulting room.

1. Capacity cost rates. The nurse works 46 weeks of 2,250 minutes, or 103,500 minutes, and practical capacity at 80% is 82,800 minutes, so a cost of GBP 49,680 gives 49,680 / 82,800 = GBP 0.60 per minute. The assistant has the same capacity at GBP 29,808, or 29,808 / 82,800 = GBP 0.36 per minute. The room offers 250 days of 480 minutes, or 120,000 minutes, of which 85% is 102,000, so GBP 20,400 a year gives 20,400 / 102,000 = GBP 0.20 per minute.

2. Time equations. Nurse time is 20 minutes, plus 10 for a new patient, plus 8 when an interpreter is needed. Observations take 10 assistant minutes, the room is occupied for the nurse time plus 5 minutes of turnaround, and consumables cost GBP 4.00 per visit.

3. Cost per visit. A follow-up costs 20 × 0.60 = 12.00 for the nurse, 10 × 0.36 = 3.60 for the assistant, 25 × 0.20 = 5.00 for the room and 4.00 for consumables, a total of GBP 24.60. The other visit types follow in the same way.

4. Reconcile with the nurse budget. The clinic sees 2,000 follow-ups, 800 new visits and 200 interpreter visits. Follow-ups use 2,000 × 20 = 40,000 nurse minutes, new visits 800 × 30 = 24,000 and interpreter visits 200 × 38 = 7,600, a total of 71,600 minutes, so 71,600 × 0.60 = GBP 42,960 is assigned to patients. The other 82,800 − 71,600 = 11,200 minutes cost 11,200 × 0.60 = GBP 6,720 and are reported as unused capacity, and 42,960 + 6,720 = 49,680 matches the budget. Utilisation is 71,600 / 82,800 = 0.865, or 86.5%.

5. Cost a pathway change. Half the follow-ups, 1,000 visits, switch to a 10-minute nurse telephone review with no assistant, room or consumables, costing 10 × 0.60 = GBP 6.00. Each switch saves 24.60 − 6.00 = GBP 18.60 at TDABC rates, or GBP 18,600 in total, made up of 10,000 nurse minutes (GBP 6,000), 10,000 assistant minutes (GBP 3,600), 25,000 room minutes (GBP 5,000) and GBP 4,000 of consumables.

6. Interpret. Only the GBP 4,000 of consumables is a cash saving unless staffing or room sessions are cut. Nurse unused capacity rises from 11,200 to 21,200 minutes, or 21,200 / 82,800 = 0.256 (25.6%) of practical capacity. The freed nurse time could absorb 10,000 / 30 = 333 extra new visits a year, and that use, or a cut in nurse hours, is what gives its GBP 6,000 real value.

TDABC costs in economic evaluation and HTA

The NICE technology appraisal manual (PMG36, section 4.4.9) calls healthcare resource groups a valuable source of information for estimating resource use and notes that their national average unit costs can reduce the need for local micro-costing. Section 4.4.10 adds that HRG data may not be appropriate when the new technology and its comparator fall under the same HRG, or when the mean cost does not reflect resource use with the new technology, and that micro-costing studies may then be more appropriate. Da Silva Etges and colleagues note that health organisations often employ TDABC in such micro-costing studies.

The English vaccination study shows the kind of output the method gives. Crocker-Buque and colleagues found a mean cost of GBP 18.20 for a childhood vaccination appointment and GBP 14.05 for an adult appointment, of which 75% was for staff, 24% for facilities and 1% for consumables. They linked lower costs to shorter allocated appointments and to healthcare assistants giving adult vaccinations, an explanation no national average gives. In a cost-utility analysis such unit costs feed the cost side of the model, and uncertain capacity and time estimates belong in the sensitivity analysis.

Boundaries with activity-based costing, micro-costing and price lists

TDABC is a type of activity-based costing that differs from its parent in the source of its rates. Conventional ABC builds activity pools from surveyed shares of time and can use many cost drivers, while TDABC uses time as the driver wherever capacity is measured in time. It belongs with bottom-up costing and contrasts with top-down costing, which divides an organisation's spending across its outputs.

The gap between a TDABC cost and an internal price list can be large. In Keel and colleagues' study of a Stockholm centre for patients with multiple chronic conditions, TDABC costs ranged from EUR 51 for a nurse telephone consultation to EUR 369 for a team visit, while the hospital's economy department price list carried EUR 187.2 for each of these activities. A flat price of that kind lets short activities subsidise long ones.

Where TDABC estimates go wrong

Most errors come from the two inputs that the method makes look simple, practical capacity and activity time. The rest come from reading a full cost as a cash effect.

1. Practical capacity set too high. Leaving out breaks, holidays and training understates every rate, as in the eight studies Shakya and colleagues identified.
2. Times taken on trust. Only 6% of the studies Shakya and colleagues reviewed used real-time tools such as time-tracking software.
3. Full cost read as a saving. Minutes removed from a pathway become cash only if spending changes.
4. Costly to scale. Keel and colleagues judged that scaling would be difficult without data infrastructures linking patients and resources across financial, clinical and process data.
5. Local results. The adult vaccination appointment cost ranged from GBP 7.59 to GBP 20.88 across the nine practices, so a TDABC cost describes the staffing and organisation of the site that produced it.

Sources

• Crocker-Buque T, Mohan K, Ramsay M, Edelstein M, Mounier-Jack S. What is the cost of delivering routine vaccinations at GP practices in England? A comparative time-driven activity-based costing analysis. Human Vaccines & Immunotherapeutics. 2019;15(12):3016-3023. https://doi.org/10.1080/21645515.2019.1619403
• da Silva Etges APB, Cruz LN, Notti RK, Neyeloff JL, Schlatter RP, Astigarraga CC, Falavigna M, Polanczyk CA. An 8-step framework for implementing time-driven activity-based costing in healthcare studies. European Journal of Health Economics. 2019;20(8):1133-1145. https://doi.org/10.1007/s10198-019-01085-8
• Kaplan RS, Anderson SR. Time-driven activity-based costing. Harvard Business Review. 2004;82(11):131-138. Excerpt published as Rethinking activity-based costing, HBS Working Knowledge, 24 January 2005. https://www.library.hbs.edu/working-knowledge/rethinking-activity-based-costing
• Kaplan RS, Porter ME. How to solve the cost crisis in health care. Harvard Business Review. 2011;89(9):46-52. https://pubmed.ncbi.nlm.nih.gov/21939127/
• Keel G, Muhammad R, Savage C, Spaak J, Gonzalez I, Lindgren P, Guttmann C, Mazzocato P. Time-driven activity-based costing for patients with multiple chronic conditions: a mixed-method study to cost care in a multidisciplinary and integrated care delivery centre at a university-affiliated tertiary teaching hospital in Stockholm, Sweden. BMJ Open. 2020;10(6):e032573. https://doi.org/10.1136/bmjopen-2019-032573
• National Institute for Health and Care Excellence. NICE technology appraisal and highly specialised technologies guidance: the manual (PMG36). Published 31 January 2022, last updated 13 October 2026. Sections 4.4.9 and 4.4.10. https://www.nice.org.uk/process/pmg36/chapter/economic-evaluation-2
• Shakya S, Bohingamu Mudiyanselage S, Robinson S, Randall S, Gao L. Time-driven activity-based costing and its use in health economic analysis: a systematic literature review. Applied Health Economics and Health Policy. 2025;23(6):989-1004. https://doi.org/10.1007/s40258-025-00988-3