1

Supplemental Table B.Summary of steps used to perform crisp-set QCA

csQCA steps / Application of QCA steps in the current study
Step 1:(a) Determine, define, and operationalize the outcome of interest
(b) Assign dichotomous set membership scores for the outcome / (a) Outcome =patient follow-through (PF); defined as the percentage of patients at each respective institution who follow-through with germline testing following a tumor screen suggestive of Lynch syndrome (LS). PF was operationalized usingordinal response options from a question on the initial survey.
(b)Possible PF-scores ranged from 1-7 and were categorized into three groups or sets: High-PF; Medium-PF; and Low-PF. Institutions with a PF score 4 were included in the High-PF set (coded as High-PF=1). All other institutions were coded as High-PF=0 and are referred to with a tilde to indicate they are not in the High-PF set (i.e., ~High-PF). For the second analysis, institutions with a PF score of 1 were coded as Low-PF=1 and all others as Low-PF=0.
Step 2: Select Cases / To maximize both sample size and diversity in implementation conditions, all institutions that met the stringent inclusion criteria (described in the manuscript) were used in the analysis.
Step 3:
(a) Identify key conditions
(b) Assign dichotomous set membership scores for each condition
(c) Create a data matrix of scores for conditions / (a) Based on theory and knowledge of UTS, the following conditions were hypothesized to be associated with High-PF when either present (+) or absent (-): 1) reflex testing on a subset of tumors is performed automatically to rule out patients with an initial positive screen who do not need genetic counseling and germline testing (+); 2) genetic counselor (GC) discloses positive screening results to patients (+); 3) difficulty contacting patients was a reported barrier (-); 4) requirement for referral from another health care provider was a reported barrier (-). Similarly, the following conditions were hypothesized to be associated with Low-PF: 1) number of challenges to adoption were to number of facilitators (+); 2) GC receives patient information on all positive screens (-); 3) GC discloses positive screening results to patients (-); 4) requirement for referral from another health care provider was reported as a barrier (+).
(b) All conditions were already dichotomized as either present=1 or absent=0 based on how they were asked in the survey.
(c) A data matrix shown in Table 2was created by listing membership scores for the outcome and key conditions for each institution.
Step 4: Determine whether conditions are necessary for the outcome / None of the conditions were originally hypothesized to be necessary for either High-PF or Low-PF. Thus, a necessary analysis was not conducted.

Supplemental Table B (continued)

Step 5: Determine whether certain conditions are sufficient for the outcome using the “truth table” approach / Even if they are not necessary for the presence of High- or Low-PF, certain conditions may still be sufficient for the respective outcome either when occurring alone or in combination with other conditions. Using fsQCA 2.0, two truth tables were created showing all possible configurations of conditions for each of the two selected outcomes (i.e., High-PF and Low-PF).
Step 6: Examine the truth table and resolve contradictions / No contradictions were identified.
Step 7: Use computer software to generate solutions through multiple comparisons of case configurations in the truth table / Using fsQCA 2.0 software, a “Standard Analysis” was performed to identify conditions associated with High-PF. A second analysis was performed to identify conditions associated with Low-PF. This software uses the Quine-McCluskey algorithm (based on Boolean simplification) to make multiple comparisons of case configurations represented in the truth table and logically simplify the data. The idea behind this minimization procedure is that if two configurations differ in only one condition, yet produce the same outcome, then the condition that distinguishes the two configurations can be considered irrelevant to the outcome and be removed to create a simpler solution.
During this process, input from the researchers was required to select prime implicants and determine which simplifying assumptions were tenable. The software then used this information to generate three solutions (complex, parsimonious, and intermediate) with High-PF as the outcome. In a separate analysis three solutions were similarly generated with Low-PF as the outcome. Only the intermediate or parsimonious solutions are shown in Table 4. The other solutions are available upon request from the first author.
Step 8: Determine if the influence of conditions is symmetrical / To determine if conditions associated with High-PF are the same as those associated with the absence of the outcome (~High-PF), steps 4-6 were repeated using ~High-PF as the outcome. Similarly, these steps were repeated using ~Low-PF as the outcome.
Step 9: Evaluate the consistency and coverage of the solutions / For each of the four outcomes analyzed (High-PF, Low-PF, ~High-PF, and ~Low-PF) the overall solution consistencies were 1; indicating that the respective combination of conditions were consistently associated with the respective outcome. For each analysis the overall coverage was 1; indicating that all of the cases with the presence of the outcome fit the solution.
Step 10: Interpret the resulting solutions and create causal models / Even when conditions are uniquely and consistently associated with an outcome, it does not necessarily mean they cause the outcome. However, these solutions in conjunction with theories, frameworks, and details about the cases were used to hypothesize amodel that describes how the conditions might lead to the outcome.Models are described in the discussion.