Patient time cost with time valued by employment status

Values a person's hours in care at the person's own hourly wage when employed for wages and at an hourly value of home production otherwise, the approach Yabroff and colleagues used in a sensitivity analysis, and multiplies by the person's hours from HE-FM-PTC-001 or HE-CF-PTC-001. The base case of the same study used one median wage for everyone, the w of HE-FM-PTC-001, so that each person's time counted equally. In that study the choice cut annual time costs for cancer survivors aged 65 and over to USD 359, against USD 913 in the base case. The value differs from the human capital value of HE-FM-ABSN-002, which values absence from paid work as lost output, usually at gross earnings including employer-paid costs.

Signature

v_i = E_i * w_i + (1 - E_i) * w_hp; C_T_i = h_i * v_i
Inputs
InputsDefinitionUnit
E_iIndicator equal to 1 when person i is employed for wages and 0 otherwiseindicator, no unit
w_iOwn hourly wage of person i when employed for wagescurrency per hour
w_hpHourly value of home production applied to people not employed for wagescurrency per hour
h_iHours person i spends travelling to, waiting for and receiving care over the periodhours per period
Output
v_iHourly value applied to the time of person icurrency per hour
C_T_iPatient time cost of person i over the periodcurrency per person per period

Function

Patient time cost function from service use, hours per use and an hourly value of time

Maps the number of times a patient uses each type of health service, the hours each use takes in travel, waiting and care, and an hourly value of time to the money value of the patient's time in care over a period, per visit and in total. It is the resource costing sum HE-FM-COST-001 with patient hours as the resource and the hourly value as the unit cost, so the package keeps only the steps particular to patient time, which are hours built up from service counts (HE-FM-PTC-001 and HE-CF-PTC-001), the change in an incremental cost-effectiveness ratio when the time cost is added (HE-FM-PTC-002) and a value of time that depends on employment status (HE-FM-PTC-003). Productivity cost is a different cost category. The absence records HE-FN-ABSN-001, HE-FM-ABSN-002 and HE-FM-ABSN-003 value time lost from paid work as lost output, under the human capital approach or within the friction period, whereas patient time cost values the hours spent in care whether they are taken from paid work, unpaid work or leisure, and an hour taken from paid work is counted under one heading only. A value of time from another price year is moved with HE-FM-COST-002, time costs in later years are discounted with HE-FN-DISC-001, and a national total from a mean time cost per patient follows HE-FM-COI-001. Notation follows the Patient Time Cost article.

Try this function

Implementations

  • Excel

    Patient time cost by employment status from named cells

    With the indicator in EmployedFlag, the own wage in OwnHourlyWage, the home production value in HomeProductionValue and the hours in PersonHours, the first formula fills PersonTimeValue and the second PersonTimeCost. Filled down one row per person, AVERAGE over the PersonTimeCost column gives the mean time cost of the group.

    =EmployedFlag*OwnHourlyWage+(1-EmployedFlag)*HomeProductionValue; =PersonHours*PersonTimeValue

Assumptions

  • Employment status held fixed over the costing period

    E_i is 1 or 0 for the whole period. In their sensitivity analysis Yabroff and colleagues used actual hourly wages for survey participants employed for wages and USD 5.32 an hour, the value of home production adjusted to 2011 dollars, for those not employed for wages. A person who works for part of the period needs the hours split by status, and a group share in place of E_i gives the bias described in HE-ER-PTC-006.

  • Wage as the proxy for the value a patient places on time

    The wage stands for the value a person places on time, which Russell argues is a reasonable proxy, adding that wage data are well measured and readily available; this formula uses a person's own wage when employed for wages, as in the sensitivity analysis of Yabroff and colleagues. Each hour is valued once whatever the patient gave up. Tranmer and colleagues review the human capital method, the friction cost method and the Washington Panel approach for patient and carer time and find the methods of valuation inconsistent, so the values chosen are stated and varied in sensitivity analysis.

Worked examples

  • Person not employed for wages valued at home production

    Illustrative hours of 55.1 a year, the adjusted mean for cancer survivors aged 65 and over in Yabroff and colleagues, for a person not employed for wages and so with no own wage, valued at USD 5.32 an hour, give USD 293.13 a year (computed here for illustration).

    E_i = 0; w_i = 0; w_hp = 5.32; h_i = 55.1; v_i = 5.32; C_T_i = 293.13
  • Employed person valued at an own hourly wage

    An employed person with 30.2 hours a year, the adjusted mean for survivors aged 18 to 64 in the same study, and an illustrative own wage of USD 22.00 an hour has a time cost of USD 664.40 a year.

    E_i = 1; w_i = 22; w_hp = 5.32; h_i = 30.2; v_i = 22; C_T_i = 664.4
  • Own wage equal to the median wage reproduces the single-wage time cost

    With an own wage equal to the study's median wage of USD 16.57 the 55.1 hours are worth USD 913.01, close to the USD 912.96 that Table 4 of the study reports for survivors aged 65 and over in the base case, the gap coming from the rounding of the hours (agreement computed here).

    E_i = 1; w_i = 16.57; w_hp = 5.32; h_i = 55.1; v_i = 16.57; C_T_i = 913.01

Common errors

  • Group means of hours and value of time multiplied in place of person-level costs

    Multiplying mean hours by a mean value of time ignores that hours and employment go together. In an illustrative group where 30% are employed, with 20 hours a year at USD 15, and 70% are not, with 50 hours at USD 5.32, the person-level mean is USD 276.20, while 41 mean hours times a mean value of USD 8.224 gives USD 337.18, 22% too high (computed here for illustration).

  • Valuation by employment status treated as neutral between groups

    The same 55.1 hours are worth USD 293.13 at the home production value and USD 913.01 at the median wage, 68% less for a person not employed (computed here for illustration). Yabroff and colleagues warn that methods which value time differently for different populations may lead to inequities, particularly for low-income and other vulnerable groups, and kept a single median wage for their base case.

Sources

  • Own wages and home production value in the sensitivity analysis of Yabroff and colleagues

    Yabroff KR, Guy GP Jr, Ekwueme DU, McNeel T, Rozjabek HM, Dowling E, Li C, Virgo KS. Annual patient time costs associated with medical care among cancer survivors in the United States. Medical Care. 2014;52(7):594-601. Methods, Measures subsection, which values patient time at the median U.S. wage of USD 16.57 per hour in 2011 in the base case and, in sensitivity analyses, at actual hourly wages for participants employed for wages and at USD 5.32 per hour, the value of home production adjusted to 2011 dollars, for those not employed; Table 4, which reports annual time costs of USD 912.96 and USD 358.80 for survivors aged 65 and over under the two approaches; and the Discussion, which explains the choice of a single median wage valuing each person's time equally and warns of inequities for low-income and other vulnerable populations.

    View source →

  • Wage as a reasonable proxy for the value patients place on their own time

    Russell LB. Completing costs: patients' time. Medical Care. 2009;47(7 Suppl 1):S89-S93. Abstract, which states that the concept of opportunity cost, often approximated by a market price, is the basis for valuing all resources, that the wage is a reasonable proxy for the value patients place on their own time, and that wage data are well measured and readily available.

    View source →

  • Review of methods for valuing patient and caregiver time by Tranmer and colleagues

    Tranmer JE, Guerriere DN, Ungar WJ, Coyte PC. Valuing patient and caregiver time: a review of the literature. PharmacoEconomics. 2005;23(5):449-459. Abstract, which presents and critiques the human capital method, the friction cost method and the Washington Panel approach, finds the methods of valuation inconsistent and frequently controversial, and notes a lack of research on time lost from unpaid labour and leisure.

    View source →

Canonical Identity

Stable URI · Machine-readable · Resolvable · CC BY 4.0