Causal analyses with target trial emulation for real-world evidence removed large self-inflicted biases: systematic bias assessment of ovarian cancer treatment effectiveness.

BACKGROUND AND OBJECTIVES: Drawing causal conclusions from real-world data (RWD) poses methodological challenges and risk of bias. We aimed to systematically assess the type and impact of potential biases that may occur when analyzing RWD using the case of progressive ovarian cancer. METHODS: We retrospectively compared overall survival with and without second-line chemotherapy (LOT2) using electronic medical records. Potential biases were determined using directed acyclic graphs. We followed a stepwise analytic approach ranging from crude analysis and multivariable-adjusted Cox model up to a full causal analysis using a marginal structural Cox model with replicates emulating a reference randomized controlled trial (RCT). To assess biases, we compared effect estimates (hazard ratios [HRs]) of each approach to the HR of the reference trial. RESULTS: The reference trial showed an HR for second line vs. delayed therapy of 1.01 (95% confidence interval [95% CI]: 0.82-1.25). The corresponding HRs from the RWD analysis ranged from 0.51 for simple baseline adjustments to 1.41 (95% CI: 1.22-1.64) accounting for immortal time bias with time-varying covariates. Causal trial emulation yielded an HR of 1.12 (95% CI: 0.96-1.28). CONCLUSION: Our study, using ovarian cancer as an example, shows the importance of a thorough causal design and analysis if one is expecting RWD to emulate clinical trial results.

[Linkage of Clinical and Claims Data in the Evaluation of Post-Stroke Care - SeDaStro: Experiences from the Tyrolian StrokeCard Program]. / Linkage von klinischen Primärdaten und Krankenkassenabrechnungsdaten in der Evaluation der Schlaganfallversorgung ­ SeDaStro: Erfahrungen aus dem Tiroler StrokeCard-Programm.

AIMS AND OBJECTIVES: Data linkage is of paramount importance in the evaluation of treatment regimens for chronic diseases where different health care sectors are involved. A comprehensive picture of long-term treatment effects and, in particular, the cost-effectiveness ratio of treatment approaches can only be drawn when data from various sources are merged and analyzed together. METHODOLOGICAL PROBLEMS AND CHALLENGES: Regarding post-acute stroke care, the present study gives an example of an exact deterministic data linkage procedure including clinical patient records and claims data of TGKK, the main Tyrolean statutory health insurance fund. Typical problems known from other data linkage projects also emerged in the so-called StrokeCard program conducted at the Medical University of Innsbruck. Distinctive Austrian features (the majority of the Austrian population benefits from a mandatory social insurance system without freedom of choice) facilitated the feasibility of the data linkage procedures. RESULTS: Over the recruitment period 01/2014-12/2015, 540 patients could be assigned to the operative dataset. Of these, 367 patients were part of the StrokeCard group (i. e. the treatment group), and 173 belonged to the usual care group (i. e. the control group); 11 patients did not complete the one-year follow-up period (7 treatment group patients vs. 4 control group patients); 7 of them died during the study (5 treatment group patients vs. 2 control group patients). For all 540 patients, TGKK claims data were available for the time-frames of one year before recruitment and one year after discharge from the University hospital. All data could be used in the health-economic evaluation of the StrokeCard program. CONCLUSIONS: The linking of clinical patient records with data collected by SHI funds opens a window of opportunities for analyses of medical care. Counter-intuitively, Austrian health services research activities have limited experience in data linkage approaches, alhough studies based on the linkage of clinical patient records and claims data are indispensable for the evaluation of complex multi-sectoral treatment schemes. The current project proves the feasibility of data linkage mechanisms in the Austrian context. This should be regarded as an impetus for extending data linkage principles to evaluation studies in the future.

Recurrence of genitals warts in pre-HPV vaccine era after laser treatment.

PURPOSE: Human papillomavirus (HPV) can cause condylomata acuminata, also known as genital warts. Our aim was to evaluate the long-term recurrence of genital warts after primary carbon dioxide laser treatment before the introduction of the vaccination against HPV. METHODS: Recurrence rate and localization of genital warts were analysed in a retrospective study in 1798 women presenting with a new diagnosis of genital warts from 1992 to 2009 at a University hospital and had received laser treatment. Additionally, data on topography, pregnancy status, and cervical smear were available for women treated from 2003 to 2009 (n = 825, data subset 1) and systematic follow-up data for women treated in 2006 and 2007 (n = 242, data subset 2). RESULTS: Median time from laser treatment to first recurrence was 14.6 weeks (data subset 2). The site most affected was the vulva (90.7%) followed by the perineum/perianal region (59.3%) and the vagina (47.3%). Abnormal Pap smear was observed in 22.6%. Systematic follow-up with a median follow-up time of 3.1 years revealed at least one recurrence in 68 (28.1%) of 242 women. Women with multifocal genital warts had a 2.9 times increased risk for recurrence compared to women with unifocal lesions (p = 0.01). CONCLUSIONS: Nearly 30% of women presenting with genital warts experienced at least one recurrence after treatment with carbon dioxide laser. Multifocal lesions are the strongest indicator of recurrence. These data provide an important insight to recurrence rates of genital warts before HPV vaccination and underline the significance of a long-term follow-up and HPV vaccination.

Short-term effects of bright light therapy in adults with chronic nonspecific back pain: a randomized controlled trial.

OBJECTIVE: The present trial evaluated incorporation of bright light therapy in the treatment of chronic nonspecific back pain (CNBP). DESIGN: A prospective, randomized, controlled, multicenter, open design with three parallel trial arms was used. SETTING: Subjects received a novel therapeutic, an expected therapeutic ineffective low dose, or no light exposure at three different medical centers. PATIENTS: A total of 125 CNBP patients reporting pain intensity of ≥3 points on item 5 of the Brief Pain Inventory (BPI) were included. INTERVENTION: Over 3 weeks, 36 active treatment, 36 placebo controls, and 33 controls received 3 or no supplementary light exposures of 5.000 lx or 230 lx, respectively. OUTCOME MEASURES: Changes in self-reported scores of pain intensity (BPI sub-score 1) and depression (Hospital Anxiety and Depression Questionnaire) were the primary outcome measures. Secondary outcome measures were changes in self-reported overall pain sensation (BPI total score), grade of everyday life impairment (BPI sub-score 2), mood (visual analog scale), and well-being (World Health Organization-Five Well-Being Index). RESULTS: Changes in pain intensity were higher (1.0 [0.8-1.6]) in the bright light group compared with controls (0.3 [-0.1-0.8]; effect size D = 0.46). Changes in the depression score were also higher in the intervention group (1.5 [0.0-2.5]) compared with controls (0.0 [0.0-2.0]; effect size D = 0.86). No differences were seen in change scores between intervention vs sham group. CONCLUSION: The present randomized controlled trial shows that light therapy even in low dose could improve depressive symptoms and reduce pain intensity in CNBP patients. Further research is needed for optimizing parameters of frequency, dose, and duration of therapeutic light exposure.