The association between drug industry payments and NCCN guideline panel membership.

Abstract

2068 Background: The high frequency of financial relationships between the pharmaceutical industry and influential oncologists who author clinical practice guidelines may influence guideline recommendations. Therefore, we assessed the financial relationships held by NCCN Guidelines panelists before and after joining the panel, compared to those held by a matched set of oncologists. Methods: Membership of NCCN Guidelines panels for the 20 most common cancers was obtained from archival guidelines and linked manually to Open Payments records of industry payments. We identified physicians who newly joined an NCCN panel during the August 2013-December 2018 study period, and we included medical oncologists who had at least 1 year of Open Payments data before and after joining. These medical oncologists who joined an NCCN panel (panelists) were matched 1:2 to medical oncologists with the same gender, institutional affiliation, and medical school graduation year, who did not join an NCCN panel (non-panelists). The dollar value of industry payments was then calculated over the 1 year before (pre-join) and after (post-join) the date that each panelist joined. We used generalized linear models to assess differences in industry payments between the panelists and matched non-panelists in the pre-join period. We used difference-in-difference estimation (DiD) to assess whether joining an NCCN panel was associated with increased payments in the post-join period. Results: There were 54 panelists and 108 non-panelists (matched from 1447 eligible oncologists at NCCN institutions). Mean per-oncologist payments among panelists were greater than non-panelists in the pre-join period ($11,259 vs $3,427, p = 0.02). From the pre-join to post-join period there was a similar increase in mean per-oncologist payments among panelists and non-panelists ($2,236 vs. $1,569, DiD estimate +$667, p = 0.77). Conclusions: Medical oncologists who were selected to an NCCN Guidelines panel had greater financial ties to industry compared to peer oncologists who were not selected. This difference was present prior to joining; oncologists did not experience a greater increase in financial payments from industry in the 1-year period after joining an NCCN panel. These results suggest an opportunity to reduce the potential influence of industry in oncology clinical practice guidelines through the selection of guideline panelists with fewer ties to industry.