Our base case consists of a Swiss patient with an oropharyngeal squamous cell carcinoma (OPSCC), age 55, with operable Tcategory (T1 or T2) OPSCC and a probability of regional disease (N+) between 60–70%. Our analyses are performed from a Swiss hospital payer perspective and with a lifetime horizon.
We developed a twostage model based on (i) a published model about the economic evaluation of TORS vs radiotherapy[13], (ii) additional literature[6] and (iii) authors’ expertise and statistics from the Centre Hospitalier XXX and University Hospital XXX.
The firststage decision tree accounts for shortterm outcomes of the surgery and its complications which are, in turn, carried forward as initial conditions for a secondstage model representing longterm outcomes through a Markov process.
The firststage decision tree accounts for shortterm outcomes of the surgery and its complications which are, in turn, carried forward as initial conditions for a secondstage model representing longterm outcomes through a Markov process.
The firststage model is depicted in Fig. 1. The two surgical strategies constitute alternatives for the first decision node, after which a chance node distinguishes between cases undergoing surgery alone, and cases requiring adjuvant radiotherapy (RT) or chemoradiotherapy (CRT). Finally, potential complications of the surgical interventions and, where appropriate, associated adjuvant (ADJ) therapy are modeled.
The secondstage model deals with longterm outcomes and is constituted by a Markov model (Fig. 1b). It represents patients entering a state of remission after treatment and models their transitions through other possible health states until death. The Markov cycle has been set to 3 months and the time horizon is the entire patient life. Initial rewards of each Markov model are carried forward from the results of the firststage model.
Model parameters representing estimates of probabilities of adjuvant treatment were derived from data published by Li et al. [6], which constitute the largest and most recent published database study with relevant outcome data. Other parameters modeling clinical events such as complication rates and recurrence rates were determined from systematic review of the literature[15]. The hospital admission rate for CRT was set at 75% of patients to be admitted once, and 25% twice. The proportion of patients needing hospital admission for RT was set at 25% only once. Regarding the need for a gastrostomy we considered a PEGrate for CRT of 70% while 20% for RT as per institutional data from CHUV and USZ.
Transition probabilities between different health states of the Markov models were directly adopted from de Almeida et al. [13] and no relevant difference in survival is assumed between the TORS and TLM arms of the model, based on a recent retrospective analysis of the National Cancer Data Base (NCDB) [6]. Table 1, probabilities of events section, reports the specific value used for each parameter. Risk of death from noncancerspecific causes is modeled following Swiss life tables, acquired through the Swiss office federal de la statistique. In order to perform probabilistic sensitivity analysis (PSA) all parameters were represented using probability distributions. Probabilities of event occurrence were represented as beta distributions, as indicated for variables ranging from 0 to 1 [16] (Table 1).
Costs were directly acquired from the “Centre hospitalier XXX” and “ University hospital XXX” administrative departments. Having adopted a hospital perspective, costs incurred by the patient are not considered in our analyses. All costs are represented as Gamma distributions, as suggested by Huinink et al [16] for values greater or equal than 0 (Table 1).
Utility coefficients (UCs) for the health states included in the model were collected with Standard Gamble method through our UceWeb [17, 18] platform from a set of 41 Swiss healthy volunteers. 17 different scenarios were evaluated by each participant [19]. Rating Scale method was also administered, to familiarize participants with the tool and as a consistency check of the obtained values. As for probabilities, UCs are represented as beta distributions (Table 1).
Table 1
Model parameters: Probabilities of occurrence of events, costs and utilities.
Variable name

Description

Mean

Standard deviation

Distribution type

Parameter 1 (alpha)

Parameter 2 (beta)

Probabilities of events

pes

Probability of esophageal stenosis

0.0476

0.0005

Beta

4

80

phem

Probability of hemorrhage

0.0243

0.0001

Beta

6

241

pho_adj

Probability of hospital readmission after adjuvant

0.1731

0.0027

Beta

9

43

pho_s

Probability of hospital readmission (TORS or TLM)

0.0333

0.0010

Beta

1

29

plg

Probability of longterm gastrostomy (1 year) after adjuvant treatment

0.0500

0.0003

Beta

9

171

plt

Probability of longterm tracheostomy (1 year)

0.0226

0.0001

Beta

4

173

psg

Probability of shortterm (6 months) gastrostomy (TORS or TLM)

0.0144

0.0001

Beta

2

137

psg_adj

Probability of shortterm (6 months) gastrostomy after adjuvant

0.2991

0.0019

Beta

32

75

por

Probability of osteoradionecrosis

0.0265

0.0002

Beta

4

147

ppf

Probability of pharyngocutaneous fistula

0.0253

0.0001

Beta

10

385

pTLMAlone

Probability of TLM alone

0.4085

0.0007

Beta

134

194

pTorsAlone

Probability of TORS alone

0.3740

0.0001

Beta

824

1379

pCRT_TLM

Probability of adjuvant CRT (TLM)

0.6289

0.0012

Beta

122

72

pCRT_tors

Probability of adjuvant CRT (TORS)

0.5272

0.0002

Beta

727

652

pRT_TLM

Probability of adjuvant RT (TLM)

0.3711

0.0012

Beta

72

122

pRT_tors

Probability of adjuvant RT (TORS)

0.4728

0.0002

Beta

652

727

plr*

Probability of local or regional recurrence (first 2 years)

0.0064

0.0000

Beta

11

1715

prr*

Probability of regional recurrence (first 2 years)

0.0064

0.0000

Beta

11

1715

pdr*

Probability of distant recurrence (first 2 years)

0.0038

0.0000

Beta

11

2900

Costs (CHF)

cTORS

Cost of TORS

14739

869.31

Gamma

287.4635

0.0195

cTLM

Cost of TLM

12671

516.23

Gamma

602.4698

0.0475

cCRT

Cost of adjuvant CRT

33911

2079.08

Gamma

266.0350

0.0078

cRT

Cost of adjuvant RT

27962

1714.35

Gamma

266.0342

0.0095

cES

Cost of esophageal stenosis

2362

410.65

Gamma

33.0832

0.0140

cGAST

Cost of gastrostomy

4332

410.65

Gamma

111.2820

0.0257

cHR_adj

Cost of hospital readmission (for adjuvant)

10097

619.05

Gamma

266.0342

0.0263

cHR_s

Cost of hospital readmission (TORS or TLM)

8203

803.41

Gamma

104.2498

0.0127

cORN

Cost of osteoradionecrosis

32111

1077.71

Gamma

887.7769

0.0276

cPF

Cost of pharyngocutaneous fistula

82892

333.96

Gamma

61609.3654

0.7432

cPH

Cost of hemorrhage (from surgical site)

4469

415.50

Gamma

115.6865

0.0259

cTRACH

Cost of tracheostomy

11688

612.67

Gamma

363.9366

0.0311

cREM

Cost of remission 02 y

168.5

9.68

Gamma

303.2588

1.7998

c2REM

Cost of remission 25 y

60

9.68

Gamma

38.4518

0.6409

cPC

Cost of palliative care

4137

367.86

Gamma

126.4754

0.0306

cRR

Cost of regional recurrence

7047

464.24

Gamma

230.4227

0.0327

cLR_chemorad

Cost of local recurrence (chemoradiation)

34041

2079.08

Gamma

268.0786

0.0079

cLR_s

Cost of local recurrence (surgical resection)

40513

2050.27

Gamma

390.4511

0.0096

cDM

Cost of distant metastasis

4137

367.86

Gamma

126.4754

0.0306

cPanendo

Cost of panendoscopy

388

23.79

Gamma

266.0337

0.6857

Utilities

uSURG

Utility coefficient of TORS or TLM

0.902

0.203

Beta

1.0328

0.1122

uRT

Utility coefficient of adjuvant RT

0.850

0.275

Beta

0.5831

0.1029

uCRT

Utility coefficient of adjuvant CRT

0.794

0.317

Beta

0.4984

0.1293

uHR

Utility coefficient of hospital readmission

0.954

0.140

Beta

1.1820

0.0570

uPF

Utility coefficient of pharyngocutaneous fistula

0.932

0.194

Beta

0.6374

0.0465

uPH

Utility coefficient of postoperative hemorrhage

0.910

0.203

Beta

0.8986

0.0889

Ug

Utility coefficient of gastrostomy

0.916

0.209

Beta

0.6975

0.0640

Ult

Utility coefficient of longterm tracheostomy

0.852

0.271

Beta

0.6109

0.1061

ues

Utility coefficient of esophageal stenosis

0.826

0.284

Beta

0.6459

0.1361

uORN

Utility coefficient of osteoradionecrosis

0.791

0.302

Beta

0.6428

0.1698

urem

Utility coefficient of remission after surgery and adjuvant

0.980

0.099

Beta

0.7702

0.0346

uremonlysurg

Utility coefficient of remission after TORS or TLM alone

0.957

0.151

Beta

0.9798

0.0200

ureg

Utility coefficient of regional recurrence

0.859

0.283

Beta

0.4401

0.0722

ulocxrt

Utility coefficient of local recurrence, RT

0.771

0.302

Beta

0.7216

0.2143

uloc

Utility coefficient of local recurrence, requiring surgery

0.755

0.316

Beta

0.6436

0.2088

udist

Utility coefficient of distant recurrence

0.213

0.336

Beta

0.2262

0.5106

upall

Utility coefficient of palliative care

0.307

0.350

Beta

0.1033

0.3816

*NOTE: for prr, plr and pdr 80% of recurrences were modeled in the first 2 years, and the remaining 20% between 2 and 5 years posttreatment (probabilities were adjusted accordingly, assuming 5% of patients have recurrences in the first 2 years6)

Willingnesstopay was set to 4000 CHF/QALM (i.e. 48000 CHF/QALY [20, 21]. Incremental cost was computed from the difference in expected cost (CHF) between TORS and TLM. Similarly, incremental utility was computed from the difference in expected utility between TORS and TLM. The incremental costutility ratio was derived taking the quotient between incremental cost and incremental utility. All costutility analyses were performed using TreeAge Pro 2019 software (Williamstown, MA, 2019).
Key model parameters were varied using oneway and twoway deterministic sensitivity analysis in order to assess their impact on the results. In particular, we explored the key role of adjuvant therapy (RT or CRT) after surgery and costs of treatment. Furthermore, all parameters affected by uncertainty were varied in probabilistic sensitivity analysis (PSA). Probabilistic sampling was performed from the distributions described above for probabilities (Beta), costs (Gamma) and utilities (Beta). PSA was performed using secondorder MonteCarlo simulations using 1000 simulations. Incremental cost and effectiveness were plotted with 95% confidence ellipsoids.