Nazanin is currently a postdoctoral research fellow at Harvard Medical School and Mass General Hospital. Nazanin received her PhD in Industrial Engineering and Operations Research from the University of Massachusetts (UMass) Amherst. She also holds an MS in Electrical and Computer Engineering from UMass Amherst. Her research focuses on disease modeling, control, and optimization. During her Ph.D., she closely collaborated with the HIV division at the CDC where she developed a sequential decision-making model to control HIV at a national level using reinforcement learning algorithms.
While at the ITA, Dr. Khatami worked with Dr. Mohammad Jalali (MJ) at ITA to develop machine learning models to address opioid crisis in the US. She also worked with Dr. Krishna Reddy at the MGH Medical Practice Evaluation Center (MPEC) to develop a tuberculosis transmission module for people with HIV.
Selected Publications
Stringfellow, Erin J; Dong, Huiru; Khatami, Seyedeh Nazanin; Lee, Hannah; Jalali, Mohammad S
In: Addiction, 2025, ISSN: 1360-0443.
@article{pmid39994821,
title = {The association between buprenorphine doses above 16 milligrams and treatment retention in a multi-payer national sample in the United States, 2014 to 2021},
author = {Erin J Stringfellow and Huiru Dong and Seyedeh Nazanin Khatami and Hannah Lee and Mohammad S Jalali},
doi = {10.1111/add.70002},
issn = {1360-0443},
year = {2025},
date = {2025-02-01},
journal = {Addiction},
abstract = {BACKGROUND AND AIMS: Buprenorphine-naloxone reduces overdose deaths in people with opioid use disorder (OUD). Treatment retention increases with higher daily doses. No national studies exist on retention's association with 24, 32 and 40 mg. This study aimed to: (1) estimate the effect on treatment retention of buprenorphine-naloxone doses between 4 and 40 mg compared with 16; and (2) compare the effect on treatment retention of 24, 32 and 40 mg doses.nnDESIGN: Observational cohort study in a national, multi-payer sample of prescription claims (IQVIA) of episodes involving buprenorphine-naloxone for OUD. Incident episodes started between 1 January 2014 and 31 March 2020, with a washout of 180 days. New episodes started with a 14+ day gap between prescriptions.nnSETTING: United States of America.nnPARTICIPANTS: The sample involved 620 229 episodes across 498 879 patients [42.3% female; mean age 37.9 (standard deviaion: 11.9)] who were dispensed prescriptions of buprenorphine-naloxone for OUD.nnMEASUREMENTS: The exposure was the maximum daily dose of buprenorphine-naloxone reached in the first 30 days of an episode, ranging from 4 to 40 mg. The outcome, treatment retention, was defined as having an active prescription at 1, 3, 6, 12, or 18 months. Covariates were age, sex, race and ethnicity, primary payer, and year of episode initiation.nnFINDINGS: Daily doses of 24, 32 and 40 mg increased retention compared with 16 mg at 1-18 months [adjusted odds ratio (aOR) range = 1.17; 95% confidence interval (CI) = 1.14, 1.20 at 18 months to 1.52 (CI = 1.49, 1.54) at 1 month, both for 24 mg]. In pairwise comparisons, 32 mg was favorable to 24 mg at 6, 12 and 18 months [aOR = 1.06 (95% CI = 1.02, 1.10) at 6 months; aOR = 1.09 (95% CI = 1.04, 1.14) at 12 months; aOR = 1.12 (95% CI = 1.06, 1.19) at 18 months], and 40 mg was favorable to 24 mg at 12 and 18 months [aOR = 1.10 (95% CI = 1.01, 1.21) at 12 months; aOR = 1.18 (95% CI = 1.06, 1.30) at 18 months].nnCONCLUSIONS: Daily buprenorphine-naloxone doses of 24 mg appear to be associated with increased treatment retention compared with 16 mg and, for 6+ month episodes, 32 and 40 mg appear to be associated with increased retention compared with 24 mg.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
BACKGROUND AND AIMS: Buprenorphine-naloxone reduces overdose deaths in people with opioid use disorder (OUD). Treatment retention increases with higher daily doses. No national studies exist on retention's association with 24, 32 and 40 mg. This study aimed to: (1) estimate the effect on treatment retention of buprenorphine-naloxone doses between 4 and 40 mg compared with 16; and (2) compare the effect on treatment retention of 24, 32 and 40 mg doses.nnDESIGN: Observational cohort study in a national, multi-payer sample of prescription claims (IQVIA) of episodes involving buprenorphine-naloxone for OUD. Incident episodes started between 1 January 2014 and 31 March 2020, with a washout of 180 days. New episodes started with a 14+ day gap between prescriptions.nnSETTING: United States of America.nnPARTICIPANTS: The sample involved 620 229 episodes across 498 879 patients [42.3% female; mean age 37.9 (standard deviaion: 11.9)] who were dispensed prescriptions of buprenorphine-naloxone for OUD.nnMEASUREMENTS: The exposure was the maximum daily dose of buprenorphine-naloxone reached in the first 30 days of an episode, ranging from 4 to 40 mg. The outcome, treatment retention, was defined as having an active prescription at 1, 3, 6, 12, or 18 months. Covariates were age, sex, race and ethnicity, primary payer, and year of episode initiation.nnFINDINGS: Daily doses of 24, 32 and 40 mg increased retention compared with 16 mg at 1-18 months [adjusted odds ratio (aOR) range = 1.17; 95% confidence interval (CI) = 1.14, 1.20 at 18 months to 1.52 (CI = 1.49, 1.54) at 1 month, both for 24 mg]. In pairwise comparisons, 32 mg was favorable to 24 mg at 6, 12 and 18 months [aOR = 1.06 (95% CI = 1.02, 1.10) at 6 months; aOR = 1.09 (95% CI = 1.04, 1.14) at 12 months; aOR = 1.12 (95% CI = 1.06, 1.19) at 18 months], and 40 mg was favorable to 24 mg at 12 and 18 months [aOR = 1.10 (95% CI = 1.01, 1.21) at 12 months; aOR = 1.18 (95% CI = 1.06, 1.30) at 18 months].nnCONCLUSIONS: Daily buprenorphine-naloxone doses of 24 mg appear to be associated with increased treatment retention compared with 16 mg and, for 6+ month episodes, 32 and 40 mg appear to be associated with increased retention compared with 24 mg.
Khatami, Seyedeh Nazanin; Gopalappa, Chaitra
Deep reinforcement learning framework for controlling infectious disease outbreaks in the context of multi-jurisdictions Journal Article
In: Math Biosci Eng, vol. 20, no. 8, pp. 14306–14326, 2023, ISSN: 1551-0018.
@article{pmid37679137,
title = {Deep reinforcement learning framework for controlling infectious disease outbreaks in the context of multi-jurisdictions},
author = {Seyedeh Nazanin Khatami and Chaitra Gopalappa},
doi = {10.3934/mbe.2023640},
issn = {1551-0018},
year = {2023},
date = {2023-06-01},
journal = {Math Biosci Eng},
volume = {20},
number = {8},
pages = {14306--14326},
abstract = {In the absence of pharmaceutical interventions, social distancing and lockdown have been key options for controlling new or reemerging respiratory infectious disease outbreaks. The timely implementation of these interventions is vital for effectively controlling and safeguarding the economy.Motivated by the COVID-19 pandemic, we evaluated whether, when, and to what level lockdowns are necessary to minimize epidemic and economic burdens of new disease outbreaks. We formulated the question as a sequential decision-making Markov Decision Process and solved it using deep Q-network algorithm. We evaluated the question under two objective functions: a 2-objective function to minimize economic burden and hospital capacity violations, suitable for diseases with severe health risks but with minimal death, and a 3-objective function that additionally minimizes the number of deaths, suitable for diseases that have high risk of mortality.A key feature of the model is that we evaluated the above questions in the context of two-geographical jurisdictions that interact through travel but make autonomous and independent decisions, evaluating under cross-jurisdictional cooperation and non-cooperation. In the 2-objective function under cross-jurisdictional cooperation, the optimal policy was to aim for shutdowns at 50 and 25% per day. Though this policy avoided hospital capacity violations, the shutdowns extended until a large proportion of the population reached herd immunity. Delays in initiating this optimal policy or non-cooperation from an outside jurisdiction required shutdowns at a higher level of 75% per day, thus adding to economic burdens. In the 3-objective function, the optimal policy under cross-jurisdictional cooperation was to aim for shutdowns of up to 75% per day to prevent deaths by reducing infected cases. This optimal policy continued for the entire duration of the simulation, suggesting that, until pharmaceutical interventions such as treatment or vaccines become available, contact reductions through physical distancing would be necessary to minimize deaths. Deviating from this policy increased the number of shutdowns and led to several deaths.In summary, we present a decision-analytic methodology for identifying optimal lockdown strategy under the context of interactions between jurisdictions that make autonomous and independent decisions. The numerical analysis outcomes are intuitive and, as expected, serve as proof of the feasibility of such a model. Our sensitivity analysis demonstrates that the optimal policy exhibits robustness to minor alterations in the transmission rate, yet shows sensitivity to more substantial deviations. This finding underscores the dynamic nature of epidemic parameters, thereby emphasizing the necessity for models trained across a diverse range of values to ensure effective policy-making.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
In the absence of pharmaceutical interventions, social distancing and lockdown have been key options for controlling new or reemerging respiratory infectious disease outbreaks. The timely implementation of these interventions is vital for effectively controlling and safeguarding the economy.Motivated by the COVID-19 pandemic, we evaluated whether, when, and to what level lockdowns are necessary to minimize epidemic and economic burdens of new disease outbreaks. We formulated the question as a sequential decision-making Markov Decision Process and solved it using deep Q-network algorithm. We evaluated the question under two objective functions: a 2-objective function to minimize economic burden and hospital capacity violations, suitable for diseases with severe health risks but with minimal death, and a 3-objective function that additionally minimizes the number of deaths, suitable for diseases that have high risk of mortality.A key feature of the model is that we evaluated the above questions in the context of two-geographical jurisdictions that interact through travel but make autonomous and independent decisions, evaluating under cross-jurisdictional cooperation and non-cooperation. In the 2-objective function under cross-jurisdictional cooperation, the optimal policy was to aim for shutdowns at 50 and 25% per day. Though this policy avoided hospital capacity violations, the shutdowns extended until a large proportion of the population reached herd immunity. Delays in initiating this optimal policy or non-cooperation from an outside jurisdiction required shutdowns at a higher level of 75% per day, thus adding to economic burdens. In the 3-objective function, the optimal policy under cross-jurisdictional cooperation was to aim for shutdowns of up to 75% per day to prevent deaths by reducing infected cases. This optimal policy continued for the entire duration of the simulation, suggesting that, until pharmaceutical interventions such as treatment or vaccines become available, contact reductions through physical distancing would be necessary to minimize deaths. Deviating from this policy increased the number of shutdowns and led to several deaths.In summary, we present a decision-analytic methodology for identifying optimal lockdown strategy under the context of interactions between jurisdictions that make autonomous and independent decisions. The numerical analysis outcomes are intuitive and, as expected, serve as proof of the feasibility of such a model. Our sensitivity analysis demonstrates that the optimal policy exhibits robustness to minor alterations in the transmission rate, yet shows sensitivity to more substantial deviations. This finding underscores the dynamic nature of epidemic parameters, thereby emphasizing the necessity for models trained across a diverse range of values to ensure effective policy-making.
Baker, Erin D.; Khatami, Seyedeh Nazanin
The levelized cost of carbon: a practical, if imperfect, method to compare CO2 abatement projects Journal Article
In: Climate Policy, vol. 19, no. 9, pp. 1132-1143, 2019.
@article{EDBaker2019,
title = {The levelized cost of carbon: a practical, if imperfect, method to compare CO2 abatement projects},
author = {Erin D. Baker and Seyedeh Nazanin Khatami},
url = {https://doi.org/10.1080/14693062.2019.1634508},
doi = {10.1080/14693062.2019.1634508},
year = {2019},
date = {2019-01-01},
journal = {Climate Policy},
volume = {19},
number = {9},
pages = {1132-1143},
publisher = {Taylor \& Francis},
keywords = {},
pubstate = {published},
tppubtype = {article}
}