About

John T. Farrar, M.D., Ph.D., is a Professor of Epidemiology in Biostatistics and Epidemiology at the Hospital of the University of Pennsylvania. He is a member of the University of Pennsylvania Cancer Center and serves as a Clinical Associate in the Department of Neurology. Additionally, he is a Senior Scholar at the Center for Clinical Epidemiology and Biostatistics and co-directs the Biostatistics Analysis Center and the Master of Science in Clinical Epidemiology Program at the Perelman School of Medicine. His research expertise encompasses pain, including the use of new pain medications, brain function in people with pain, complementary and alternative therapies, and new methodologies for understanding patient pain reports in clinical trials. His current research includes studies on acupuncture for osteoarthritis pain, brain imaging of pain patients, and cancer treatment-related pain. Clinically, he focuses on all aspects of pain and symptom therapy in cancer patients, contributing to the development of multidisciplinary programs for pain evaluation and treatment.

Research topics

• Medicine
• Internal medicine
• Surgery
• Psychiatry
• Physical therapy
• Anesthesia
• Medical physics
• Pharmacology
• Intensive care medicine

Selected publications

• Research recommendations for the HEAL Initiative: A path forward for pain research

  Journal of Pain · 2026-04-01

  articleOpen access

  Chronic pain conditions affect 24% of the US population and account for the greatest cause of disability, leading to tremendous suffering and lost productivity. The enormity of the problem is magnified by the dearth of safe, effective medications. We need more research that advances our understanding of pain to aid in the development of new therapies. Existing non-drug treatments are greatly underutilized for pain management despite evidence of their effectiveness, demonstrating the need for research on how best to implement these therapies. The NIH Helping End Addiction Long-term® Initiative (the NIH HEAL Initiative®) was launched in response to the opioid overdose crisis and set out to increase research and improve treatments for addiction disorders and chronic pain conditions. The HEAL Initiative® has made tremendous strides toward these goals since its initial launch in 2018. In 2024, the NIH convened a working group of external experts to assess its progress and strategize for the next five years of HEAL funding specifically for pain research. That process culminated in the production of an accepted Report containing these ten Research Priorities and five Core Principles to guide NIH leadership in planning and funding pain research within the HEAL initiative over the next five years. Here, we present these recommendations for consideration by the wider pain research community and invite further active discussion. PERSPECTIVE: Chronic pain remains a major public health crisis in the U.S. that has been insufficiently addressed. The research priorities outlined here were created to build the HEAL Initiative's pain research portfolio over the next five years. The primary aims are to develop and advance treatments for chronic pain.

  PublisherOA PDFDOI

• The impact of opioid withdrawal symptoms on pain outcomes in enriched enrollment randomized withdrawal trials: a mediation meta-analysis of trials submitted to the US Food and Drug Administration

  Pain · 2025-12-24

  articleSenior author

  ABSTRACT: The validity of enriched enrollment randomized withdrawal (EERW) trials to evaluate the efficacy of opioids for the treatment of chronic pain has been questioned. Enriched enrollment randomized withdrawal trials include an open-label titration phase to identify treatment responders who tolerate the drug, followed by a double-blind randomized phase in which responders either continue the drug or switch to placebo. A key concern is that the apparent efficacy of opioids in EERW trials may be attributable to induction of withdrawal symptoms among participants switched to placebo. We used individual participant data from 13 EERW trials (N = 5070) submitted to the US Food and Drug Administration (FDA) to estimate the extent to which withdrawal symptoms mediated the treatment effect of opioids on pain. The primary mediator was the maximum Subjective Opioid Withdrawal Scale score during the randomized phase. The primary outcome was the change in pain intensity (numeric rating scale) from randomization baseline to week 12. The pooled average treatment effect was -0.71 (95% confidence interval [CI], -0.87 to -0.55) on the numeric rating scale. Withdrawal symptoms did not significantly mediate the effect of opioids on pain overall, accounting for 2% of the pooled treatment effect (95% CI, -1% to 4%). However, significant mediation was observed in 3 individual trials (range, 8% to 28% of treatment effect mediated). Although withdrawal symptoms did not systematically bias efficacy findings in EERW trials of opioids submitted to the FDA, they contributed to overestimation in some cases. These findings support incorporating mediation analyses in future EERW trials to ensure accurate interpretation of study results.

  PublisherDOI

• Risk Factors for Fatigue in Adults Receiving Maintenance Hemodialysis Who Have Chronic Pain: A Secondary Analysis of the HOPE Consortium Trial

  Kidney Medicine · 2025-12-15

  articleOpen access

  Rationale & Objective: Fatigue is commonly experienced by adults with kidney failure receiving hemodialysis and those with chronic pain, but factors associated with fatigue are not fully understood. We determined the prevalence of fatigue in a clinical trial cohort of adults receiving maintenance hemodialysis who have chronic pain and identified factors associated with fatigue. Study Design: A cross-sectional study. Setting & Participants: The baseline data from the HOPE Consortium Trial to Reduce Pain and Opioid Use in Hemodialysis (HOPE Trial). Of the 643 participants randomized in the HOPE Trial, 636 had a baseline fatigue assessment and were included in this study. Exposures: Pain, sociodemographic, biological, dialysis-related, medical comorbid condition, psychological, and behavioral factors. Outcome: Fatigue was evaluated with the patient-reported outcomes measurement information system Fatigue SF 6a and defined as a T-score of ≥ 55. Analytical Approach: Logistic regression models. Results: Seventy-three percent of participants reported fatigue (n = 463), mean age was 60.4 (12.5), 289 (45.4%) were female, and 294 (46.2%) were Black/African American. In fully adjusted models, higher pain interference and opioid use in the last 14 days were each associated with higher odds of having fatigue (odds ratio ([OR) ] 1.37; 95% CI, 1.18-1.61; OR 1.80; 95% CI, 1.03-3.21, respectively), as were greater depressive symptoms and sleep disturbance (OR 1.21; 95% CI. 1.13-1.31; OR 1.08 95% CI 1.03-1.12, respectively). Higher physical function was associated with lower odds of having fatigue (OR 0.96 95% CI 0.93-0.99). Limitations: Fatigue assessed at one point in time. Conclusions: In adults receiving maintenance hemodialysis who have chronic pain, pain interference, opioid use, depression, and sleep disturbances are associated with increased odds of fatigue, and greater physical function is associated with lower odds of fatigue. Future work is needed to evaluate longitudinal associations, underlying mechanisms, and identify interventions.

  PublisherOA PDFDOI

• Predictors of response to full agonist opioids in enriched enrollment randomized withdrawal clinical trials: a participant-level data meta-analysis

  Pain · 2025-09-02 · 1 citations

  articleSenior authorCorresponding

  ABSTRACT: This study aims to identify predictors of success in treating chronic pain patients with full agonist opioids by analyzing harmonized individual patient data from 5594 participants in 9 enriched enrollment randomized withdrawal clinical trials available in the Food and Drug Administration data repository. We analyzed both the participants' success with titration and continued success in the 84-day maintenance phases after randomization for those maintained on the drug. We used the full data set to assess participant demographics and subsets of data containing participant reported outcomes at baseline. Participants had an average age of 51, with 55% female participants and 66% non-Hispanic white. No clinically relevant differences were observed between participants who failed titration or those who continued on full agonists through the maintenance phase. Prediction models were developed using mixed effects logistic regression and generalized linear mixed models, with the study as a random effect to account for inter-study differences. Despite large numbers, the analysis did not reveal clinically useful prediction models for either the titration or maintenance phase; however, higher initial pain scores were modest predictors of poorer outcomes. No patient-reported outcome measures were predictive of responses to therapy. The study's limitations include its volunteer-based sample and the exclusion criteria, although excluding patients with opioid use disorder or serious psychological conditions are similar to those used in clinical care. As no strong predictive factors for successful treatment were identified, the decision to use opioids to treat chronic pain requires careful clinical judgment and close monitoring.

  PublisherDOI

• Rate of biopsy prior to resection among patients with high-grade glioma: a nationwide database analysis

  Journal of neurosurgery · 2025-11-08

  article

  OBJECTIVE: In cases of high-grade glioma that ultimately proceed to resection, a purely diagnostic biopsy might be unnecessary and add undue risk and cost. The frequency of resection following biopsy in a short time frame has not been reported among patients with high-grade glioma using a nationwide database. The objective of this study was to determine the occurrence rate of resection within 60 days of biopsy among patients with high-grade glioma and to determine if this occurrence is associated with patient or hospital characteristics. METHODS: This retrospective cohort study used a large private health insurance database to identify adult patients who underwent craniotomy for resection of high-grade glioma and received radiation treatment and temozolomide within 90 days of surgery from 2006 to 2022. This patient cohort was also queried for patients who underwent surgery for biopsy within the 60 days preceding craniotomy. The patient and hospital-related factors associated with biopsy within 60 days prior to resection were then examined. RESULTS: A total of 3051 patients who underwent resection of high-grade glioma were included; 106 patients (3.5%) underwent additional surgery for biopsy within 60 days prior to resection. The likelihood of undergoing additional surgery for biopsy prior to resection decreased with increasing age (in 5-year increments, OR 0.93 [95% CI 0.87-0.99]; p = 0.04). Patients who underwent biopsy were significantly younger (p = 0.02). The Northeast region had the highest rate of preresection biopsies (4.1%, 11/267). Most patients (93.1%, 94/101) underwent preresection biopsy at non-high-volume centers, and 24.5% (23/94) of these patients subsequently underwent resection at a high-volume center. CONCLUSIONS: A low proportion of patients with high-grade glioma underwent additional surgery for biopsy prior to resection. This finding provides a baseline as preresection biopsy might become more common in the context of investigational drug development requiring pretreatment baseline specimens.

  PublisherDOI

• Predictors of successful initiation of buprenorphine in enriched enrollment randomized withdrawal clinical trials in both opioid experienced and naïve participants: a participant-level meta-analysis

  PAIN Reports · 2025-06-05

  articleOpen accessSenior authorCorresponding

  Introduction: No prediction models exist for the success for buprenorphine initiation in opioid-naïve patients or in transition from other opioids in patients treated for chronic pain. Objectives: To create a prediction model for the successful use of buprenorphine to treat chronic pain. Methods: Stepwise Akaike information criterion prediction modeling procedures were applied to a harmonized participant-level data set of 10 enriched enrollment randomized withdrawal clinical trials of buprenorphine submitted to the Food and Drug Administration. Available baseline factors and nine patient-reported outcomes were considered to predict success with the titration (10 studies) and maintenance of benefit after randomization (5 studies). Patient-reported outcomes were modeled separately given inconsistent use across studies. Results: No prediction model reached an area under the receiver operator curve ≥0.70, the threshold for clinical usefulness. Successful initiation or transition of buprenorphine was accomplished in 3541 of 6052 (58.7%) participants, and 614 of 877 (70.0%) completed the 12-week maintenance phase with no difference between opioid-experienced and opioid-naïve participants. Only a medical history of obesity and baseline pain were retained in the overall titration model and only baseline pain in the maintenance model. Only brief pain inventory and subject opioid withdrawal scores were retained in the titration subsets containing those measures. Conclusion: No clinically useful prediction models of clinical benefit were identified, but a few covariates may be of interest in future studies of the initiation of buprenorphine in opioid-naïve patients or of transition from other opioids to buprenorphine. The lack of a predictor supports considering a trial of buprenorphine in clinically relevant scenarios for patients without known opioid use disorder, including careful monitoring and an a priori plan to deal with any problems that may occur.

  PublisherDOI

• Minimum clinically important differences in acute pain: a patient-level re-analysis of randomized controlled analgesic trials submitted to the US Food and Drug Administration

  Pain · 2025-05-09 · 8 citations

  articleOpen accessSenior author

  ABSTRACT: The lack of established minimum clinically important differences in acute pain has made it challenging to interpret efficacy in analgesic trials. We performed a patient-level re-analysis of double-blind, placebo-controlled trials submitted to the US Food and Drug Administration to estimate minimum clinically important differences in acute postoperative pain. Trials were categorized by acute surgical pain model: dental extraction, bunionectomy, orthopedic surgery, and soft tissue surgery. Pain intensity was assessed using the 0 to 10 numeric rating scale (NRS) or 0 to 100 visual analog scale, with visual analog scale scores converted to NRS for analysis. To avoid misclassification from arbitrary thresholds on global impression of change or pain relief scales, meaningful pain relief was determined using the double-stopwatch technique, where patients actively indicated the times they experienced perceptible and meaningful relief. Across 29 trials, 9047 patients with moderate-to-severe baseline pain were included. Patients with severe baseline pain (NRS ≥7) reported meaningful relief at a higher absolute NRS and required larger absolute reductions in pain intensity than those with moderate baseline pain (NRS 4-<7). However, the percent reduction in pain at meaningful relief remained stable across baseline pain levels, suggesting patients assess meaningful relief in relative rather than absolute terms. No appreciable differences in the changes in pain at meaningful relief were observed by age, sex, drug, or route of administration. Receiver operating characteristic curve analysis identified a 50% reduction in pain intensity as a consistent and clinically meaningful threshold across surgical pain models, supporting its use as a standardized patient-centric metric for evaluating analgesic efficacy.

  PublisherOA PDFDOI

• Perception of neurosurgery among surgical patients with essential tremor: A qualitative mixed methods study

  World Neurosurgery X · 2025-08-11

  articleOpen access

  There is a dearth of evidence on knowledge and perceptions of procedures among patients with essential tremor (ET). The objective of this study was to utilize a mixed methods design incorporating in-depth individual interviews to investigate the perception of procedures among patients with ET who underwent surgical intervention. Semi-structured, in-depth individual interviews paired with survey questionnaires were conducted among participants with ET who had a prior surgical procedure for the disorder. Thematic analysis of qualitative data was conducted using an approach based on grounded theory methodology. Of the 20 patients interviewed, nine patients (45 %) had undergone magnetic resonance-guided focused ultrasound (MRgFUS) thalamotomy, nine patients (45 %) had undergone deep brain stimulation (DBS) implantation, and two patients (10 %) had undergone both DBS implantation and MRgFUS thalamotomy. In ranking factors from most to least important in deciding which type of surgery to undergo, patients most frequently selected safety as the most important factor (9/20, 45 %). Hair shave required was most frequently selected as the least important factor (14/20, 70 %). Seven patients (35 %) reported having zero or minimal knowledge of the risks and benefits of either MRgFUS thalamotomy or DBS before their surgery. Patients discussed their surgical outcomes including adverse effects of surgery. In deciding which type of surgery to undergo for tremor, participants discussed the role of safety, perceived invasiveness, and follow-up care required. Participants reflected on the life-changing benefits of tremor control but also discussed detrimental adverse effects such as dysarthria and gait instability following surgery.

  PublisherDOI

• Group Response Analysis: Clinically Interpretable Longitudinal Responder Analysis Methods Developed Using FDA Data

  medRxiv · 2025-08-15

  preprintOpen accessSenior author

  1 Abstract Responder analyses for the evaluation of randomized clinical trial (RCT) data have become more common in the recent past, since they can provide the medical community with results that are more directly applicable to clinical care. For pain studies, the predominant responder analysis compares the change in the individual participants’ pain level at baseline to their value at the end of the study period and uses a predetermined clinically important change cut-off value to define a response. While useful, this method substantially reduces the efficiency of the RCT by dichotomizing the results and is limited to comparing the baseline to the end of the study only. In this paper, we introduce a novel approach to the patient response over time with a focus on single dose post-operative studies. This technique provides graphical presentations and statistical approaches to understand the onset of any specified level of response, the maximum proportion of patients with a response at any point in time, and the duration of that response over time. In addition, each outcome can be summarized to examine the result across all possible cut-off points for clinically important differences (CID). We accomplish this by introducing three interrelated, longitudinal efficacy statistics: ROOT, GRO, and GROOT. The response outcome over time (ROOT) estimates the total proportion of a study period an individual patient spends as a responder. The group response outcome (GRO) estimates the instantaneous proportion of responders at all time points across the study period. The group response outcome over time (GROOT) summarizes total efficacy in a cohort, and can be calculated as the area under the GRO curve, or as the mean ROOT; they are identical. This novel method provides a clinically interpretable responder analysis over the full period of the study and, by using every data point across time, mitigates the loss of statistical power typically associated with dichotomized responder outcomes. Group response analysis is based upon repeated assessments of categorical or continuous measures categorizing each participant’s status as a treatment responder or non-responder at every timepoint based on the prespecified clinically important difference. Both the visual and statistical comparison of any two or more curves provide a comparison of the overall efficacy, which can be statistically tested using a standard asymptotic hypothesis test (such as Wald (Johnson &amp; Romer, 2016)). The method allows for an integrated evaluation of three main components of drug efficacy: the proportion of participants achieving a CID over time (effect), the time to achieve that response (onset), and the length of the response (duration). In this paper, we present the group response analysis methodology and then illustrate it using data from a placebo-controlled randomized clinical trial (RCT) for postoperative pain after third molar extraction treated with meloxicam and ibuprofen as an active comparator (Christensen et al., 2018). Our approach yields similar effect sizes as the sum of pain intensity differences (SPID) commonly used for pain study analyses while providing superior clinical interpretability and a more complete evaluation of drug therapies beyond just efficacy. We propose that this method can be used as a primary or secondary analysis of pain RCTs to answer the question of the patient response to treatment and provide suitable data to compare efficacies across treatment groups.

  PublisherOA PDFDOI

• Predictors of Supplemental Opioid Use After Third Molar Extraction

  medRxiv · 2025-07-18

  preprintOpen access

  Objectives: Non-steroidal anti-inflammatory drugs (NSAIDs) are recommended as first-line analgesics following third molar extraction, but some patients require supplemental opioids for pain management. The objective of this study was to identify demographic and clinical factors that predicted supplemental opioid use following third molar extraction in patients treated with an evidence-based analgesic regimen. Methods: Healthy adults underwent surgical extraction of partial or full bony impacted mandibular third molar. When pain intensity was ≥4/10, participants were given ibuprofen 400 mg (N=59) or placebo (N=26) in a randomized, double-blind design. After 4h, all participants transitioned to open-label ibuprofen 400 mg + acetaminophen 500 mg, with oxycodone 5 mg available for breakthrough pain. Analgesic use was documented for the first week after extraction. Predictors of supplemental opioid use in addition to ibuprofen + acetaminophen were evaluated by logistic regression. Results: Ibuprofen + acetaminophen provided adequate analgesia in most of the 85 participants, with 17 participants (20%) using supplemental oxycodone in the first week after extraction. Female sex (OR: 6.770; 95% CI: 1.657-35.57; p=0.013) and higher body mass index (BMI) (OR: 1.253; 95% CI: 1.052-1.525; p=0.016) were associated with increased odds of supplemental opioid use, while higher difficulty index (Pederson score) slightly decreased the odds of supplemental opioid use (OR: 0.852; 95% CI: 0.724-0.993; p=0.043). Adding pre-surgery neutrophil counts improved model fit, with higher neutrophil counts associated with lower odds of supplemental opioid use (OR: 0.435; 95% CI: 0.212-0.775; p=0.011). Conclusions: Female sex, higher BMI, and pre-surgery neutrophil counts were predictors of supplemental opioid use in patients treated with an evidence-based analgesic regimen. Greater surgical difficulty of third molar extraction does not increase the likelihood of supplemental opioid use.

  PublisherOA PDFDOI

Recent grants

• NIH Grant R01AT000304

  NIH · $3.4M · 2007

• NIH Grant R01CA073797

  NIH · $729k · 2003

• NIH Grant R21AT000960

  NIH · $396k · 2007

• NIH Grant K08NS001865

  NIH · $486k · 2000

Frequent coauthors

• Robert H. Dworkin

  University of Rochester Medical Center

  145 shared

• Dennis C. Turk

  Cornell University

  95 shared

• Nathaniel P. Katz

  Tufts University

  85 shared

• Raashid Luqmani

  NIHR Oxford Musculoskeletal Biomedical Research Centre

  82 shared

• David L. Hawksworth

  81 shared

• P. W. James

  University of California, Davis

  81 shared

• Bowen Huang

  81 shared

• D. H. S. Richardson

  81 shared