. 2019 Jul 8;179(9):1201–1208. doi: 10.1001/jamainternmed.2019.1678

Table 1. Variables for Base Case and Sensitivity Analyses.

Variable	Base Case	One-way Sensitivity Analysis^a		Probabilistic Sensitivity Analysis^b
Variable	Base Case	Low Value	High Value	SD	Distribution
Costs
Intermediate costs, 2017 US $^c
3-mo OCP supply	38.48	0.01	5000.00	100.68	Gamma	VA data
Dispensing for 3-mo OCP supply	8.58	0.01	110.00	6.70	Gamma	VA data
Copayment for 3-mo OCP supply	−24.00	NA	NA	NA	NA	US Department of VA,²¹ 2018
Outcome costs, $
Live birth
Prenatal care^d	2968.79	100.00	15 000.00	3789.39	Gamma	National Partnership for Women and Families²⁷ 2013
Intrapartum/delivery care	7933.67	1000.00	100 000.00	6827.42	Gamma	VA data
Newborn care (7 d)	6480.85	300.00	100 000.00	6000.00	Gamma	VA data
Miscarriage	1186.41	100.00	5000.00	1452.71	Gamma	VA data
Abortion	0	NA	NA	NA	NA	US Department of VA²⁸
Probabilities
Choose 12-mo supply	0.50	0	1.00	0.20	Beta	Assumption
Have copayments for OCPs	0.35	0	1.00	0.10	Beta	VA data
OCP use probabilities
Discontinue OCP use given 3 or 12 mo	0.35	0	0.07	0.10	Beta	Borrero et al,¹⁵2013; assumption
≥1 gap in coverage given 3 mo	0.43	0	0.70	0.15	Beta	Borrero et al,¹⁵ 2013
Annual pregnancy probabilities^e
Pregnancy given continuous OCP use	0.09	0	0.20	0.05	Beta	Trussell,⁶ 2011
Pregnancy given discontinued OCPs^f	0.47	0	1.00	0.15	Beta	Borrero et al,⁴ 2017; Trussell,⁶ 2011
Pregnancy during OCP coverage gap(s)	0.85	0	1.00	0.08	Beta	Trussell,⁶ 2011; assumption
Pregnancy outcome probabilities
Miscarriage	0.10	0	0.40	0.07	Beta	American College of Obstetricians and Gynecologists,²⁹ 2018
Abortion	0.42	0	0.70	0.05	Beta	Finer and Zolna,²⁶ 2016
Live birth	0.48	NA	NA	NA	NA	NA
Pregnancy paid for by VA	0.52	0	1.00	0.20	Beta	Borrero et al,⁴ 2017; assumption
Newborn care paid for by VA^g	0.58	0	1.00	0.20	Beta	VA data
Durations
OCP use given ≥1 coverage gap, mo	10	3	11.75^h	1.00	Beta	Borrero et al,¹⁵ 2013
OCP use given discontinuation of OCPs, mo	8	0	12.00	1.5.0	Beta	Borrero et al,¹⁵ 2013

Abbreviations: NA, not applicable; OCP, oral contraceptive pill; VA, Veterans Affairs.

^{^a}

Cost ranges for 1-way sensitivity analysis are based on extremes of empirical VA data. All other ranges are based on published literature and are intentionally broad to reflect parameter uncertainty and allow for extreme value analysis.³⁰

^{^b}

Distribution means are equal to the base case. Distributions were chosen based on established best practices.³⁰

^{^c}

Intermediate costs are scaled directly based on the minimum number of 3-month supplies necessary to account for time covered by OCPs or multiplied by 4 for 12-month supplies.

^{^d}

Prenatal care costs are based on mean costs paid by commercial insurers for prenatal care in 2010 ($2641), adjusted to 2017 US dollars.

^{^e}

Based on 1 full year of use and prorated in the model to account for the proportion of time in a given state (covered by OCPs, OCP use gap, or discontinuation of OCP use), and for a time horizon of twelve 28-day periods (336 days) instead of a full calendar year.

^{^f}

The base case probability is a weighted mean of annual typical use failure rates for nonprescription contraceptive methods (male condoms, withdrawal, fertility awareness–based methods, and spermicides) and no method use, based on the distribution of current use of these methods observed among women veterans at risk for unintended pregnancy in the Examining Contraceptive Use and Unmet Need Among Women Veterans (ECUUN) study.^4,6

^{^g}

Indicates the probability that newborn care is paid for by the VA among women whose pregnancy care is covered by the VA.

^{^h}

Indicates the maximum amount of time that can be covered by OCPs if a woman misses at least 7 days between refills in a time horizon of twelve 28-day time periods (336 days).