Skip to main content
. Author manuscript; available in PMC: 2017 Sep 1.
Published in final edited form as: J Public Health Manag Pract. 2016 Sep-Oct;22(5):452–460. doi: 10.1097/PHH.0000000000000349

Table 3. Cost per Incident Cancer Case Reported, Adjusted for Regional Cost of Living, National Program of Cancer Registries.

Dependent Variable Cost per Case—All Activitiesa,b,c

Independent Variablesd (Comparator in Italics) 95% Confidence Interval

Estimate Lower Higher
Case volume (reference: low)
 Medium case volume −0.782e −0.853 −0.711
 High case volume −1.073e −1.202 −0.944
Consolidation effort (reference: low)
 Medium consolidation effort −0.029 −0.161 0.220
 High consolidation effort 0.275f 0.019 0.531
Region (reference: South)
 Northeast −0.086 −0.309 0.137
 Midwest −0.552e −0.656 −0.449
 West −0.357e −0.444 −0.271
Geographic area (reference: small)
 Medium geographic area 0.020 −0.087 0.126
 Large geographic area 0.366e 0.189 0.542
 % of records passing 100% edits −0.098e −0.105 −0.092
 % of records submitted electronically −0.493e −0.773 −0.213
Constant 4.963
Number of observations 120.00

Abbreviation: NPCR, National Program of Cancer Registries.

a

Excludes the Pacific Regional Central Cancer Registry. Registry did not report cases diagnosed in 2006 and is still ramping up case reporting. Low case volume distorts cost per case calculations.

b

Excludes NPCR registries that also receive Surveillance, Epidemiology, and End Results funding: California (San Francisco-Oakland, San Jose-Monterey, Los Angeles), Georgia (Atlanta), Kentucky, Louisiana, Michigan (Detroit), New Jersey, and Washington (Seattle-Puget Sound).

c

Cost per case adjusted using area wage index as noted in text previously. Log values are used.

d

Final model presented previously. R2 was 0.60 for this model. The following variables were dropped from the specification because of limited explanatory power or correlation between regressors: population size flags (low, medium, high); contractor flag; and North American Association of Central Cancer Registries status (gold or silver).

e

.001%; robust standard errors.

f

.05%; robust standard errors.