references.bib

@Article{Do2019,
  title = {Red and Orange Flags for Secondary Headaches in Clinical Practice: {{SNNOOP10}} List},
  shorttitle = {Red and Orange Flags for Secondary Headaches in Clinical Practice},
  author = {Thien Phu Do and Angelique Remmers and Henrik Winther Schytz and Christoph Schankin and Sarah E. Nelson and Mark Obermann and Jakob M{\o}ller Hansen and Alexandra J. Sinclair and Andreas R. Gantenbein and Guus G. Schoonman},
  year = {2019},
  month = {jan},
  volume = {92},
  pages = {134--144},
  issn = {1526-632X},
  doi = {10.1212/WNL.0000000000006697},
  abstract = {A minority of headache patients have a secondary headache disorder. The medical literature presents and promotes red flags to increase the likelihood of identifying a secondary etiology. In this review, we aim to discuss the incidence and prevalence of secondary headaches as well as the data on sensitivity, specificity, and predictive value of red flags for secondary headaches. We review the following red flags: (1) systemic symptoms including fever; (2) neoplasm history; (3) neurologic deficit (including decreased consciousness); (4) sudden or abrupt onset; (5) older age (onset after 65 years); (6) pattern change or recent onset of new headache; (7) positional headache; (8) precipitated by sneezing, coughing, or exercise; (9) papilledema; (10) progressive headache and atypical presentations; (11) pregnancy or puerperium; (12) painful eye with autonomic features; (13) posttraumatic onset of headache; (14) pathology of the immune system such as HIV; (15) painkiller overuse or new drug at onset of headache. Using the systematic SNNOOP10 list to screen new headache patients will presumably increase the likelihood of detecting a secondary cause. The lack of prospective epidemiologic studies on red flags and the low incidence of many secondary headaches leave many questions unanswered and call for large prospective studies. A validated screening tool could reduce unneeded neuroimaging and costs.},
  journal = {Neurology},
  keywords = {Aged,Aged; 80 and over,Female,Headache,Humans,Male,Neoplasms,Nervous System Diseases,Neuroimaging},
  language = {eng},
  number = {3},
  pmcid = {PMC6340385},
  pmid = {30587518},
}
@Misc{RCoreTeam2020,
  title = {R: {{A Language}} and {{Environment}} for {{Statistical Computing}}},
  author = {{R Core Team}},
  year = {2020},
  address = {{Vienna, Austria}},
  howpublished = {R Foundation for Statistical Computing},
}
@Misc{DanielD.Sjoberg2020,
  title = {Gtsummary: {{Presentation}}-{{Ready Data Summary}} and {{Analytic Result Tables}}},
  author = {{Daniel D. Sjoberg} and {Michael Curry} and {Margie Hannum} and {Karissa Whiting} and {Emily C. Zabor}},
  year = {2020},
}

@Article{Wickham2019,
  title = {Welcome to the {{Tidyverse}}},
  author = {Hadley Wickham and Mara Averick and Jennifer Bryan and Winston Chang and Lucy McGowan and Romain Fran{\c c}ois and Garrett Grolemund and Alex Hayes and Lionel Henry and Jim Hester and Max Kuhn and Thomas Pedersen and Evan Miller and Stephan Bache and Kirill M{\"u}ller and Jeroen Ooms and David Robinson and Dana Seidel and Vitalie Spinu and Kohske Takahashi and Davis Vaughan and Claus Wilke and Kara Woo and Hiroaki Yutani},
  year = {2019},
  month = {nov},
  volume = {4},
  pages = {1686},
  issn = {2475-9066},
  doi = {10.21105/joss.01686},
  file = {/Users/danielle/Zotero/storage/F56FT56J/Wickham et al. - 2019 - Welcome to the Tidyverse.pdf},
  journal = {Journal of Open Source Software},
  number = {43},
}
@Misc{DanielDSjoberg2020,
  title = {Gtsummary: {{Presentation}}-{{Ready Data Summary}} and {{Analytic Result Tables}}},
  author = {{Daniel D Sjoberg} and {Michael Curry} and {Margie Hannum} and {Karissa Whiting} and {Emily C Zabor}},
  year = {2020},
}
@Article{AndyLiaw2002,
  title = {Classification and {{Regression}} by {{randomForest}}},
  author = {{Andy Liaw} and {Matthew Wiener}},
  year = {2002},
  volume = {2},
  pages = {18--22},
  journal = {R News},
  number = {3},
}

@Article{Breiman2001,
  title = {Random {{Forests}}},
  author = {Leo Breiman},
  year = {2001},
  volume = {45},
  pages = {5--32},
  issn = {08856125},
  doi = {10.1023/A:1010933404324},
  file = {/Users/danielle/Zotero/storage/JFD67DXR/Breiman - 2001 - [No title found].pdf},
  journal = {Machine Learning},
  number = {1},
}
@Article{Deng2014,
  title = {Interpreting {{Tree Ensembles}} with {{inTrees}}},
  author = {Houtao Deng},
  year = {2014},
  month = {aug},
  abstract = {Tree ensembles such as random forests and boosted trees are accurate but difficult to understand, debug and deploy. In this work, we provide the inTrees (interpretable trees) framework that extracts, measures, prunes and selects rules from a tree ensemble, and calculates frequent variable interactions. An rule-based learner, referred to as the simplified tree ensemble learner (STEL), can also be formed and used for future prediction. The inTrees framework can applied to both classification and regression problems, and is applicable to many types of tree ensembles, e.g., random forests, regularized random forests, and boosted trees. We implemented the inTrees algorithms in the {"}inTrees{"} R package.},
  archiveprefix = {arXiv},
  eprint = {1408.5456},
  eprinttype = {arxiv},
  file = {/Users/danielle/Zotero/storage/KSHCKUHX/Deng - 2014 - Interpreting Tree Ensembles with inTrees.pdf;/Users/danielle/Zotero/storage/6CDQ9F8L/1408.html},
  journal = {arXiv:1408.5456 [cs, stat]},
  keywords = {Computer Science - Machine Learning,Statistics - Machine Learning},
  primaryclass = {cs, stat},
}
@Article{Cain2018,
  title = {Emergency {{Department Use}} of {{Neuroimaging}} in {{Children}} and {{Adolescents Presenting}} with {{Headache}}},
  author = {Meghan R. Cain and Dimitrios Arkilo and Amy M. Linabery and Anupam B. Kharbanda},
  year = {2018},
  month = {oct},
  volume = {201},
  pages = {196--201},
  issn = {1097-6833},
  doi = {10.1016/j.jpeds.2018.05.023},
  abstract = {OBJECTIVES: To evaluate emergency department use and outcomes of neuroimaging for headache in a free-standing children's hospital system. STUDY DESIGN: We prospectively enrolled children aged 6-18 years who presented to the emergency department with a chief complaint of headache from September 2015 to September 2016. Standardized data collection was performed in real time, including telephone follow-up as needed, and imaging outcome was determined through a chart review. Using multivariable logistic regression, we estimated the associations between clinically important patient characteristics and neuroimaging. RESULTS: Of 294 enrolled patients, 53 (18\%) underwent neuroimaging (computed tomography or magnetic resonance imaging) and 2 (0.7\%) had clinically important intracranial findings. Presenting with abnormal neurologic examination findings (OR,\,11.55; 95\% CI, 3.24-41.22), no history of similar headaches (OR,\,2.13; 95\% CI, 1.08-4.18), and white race (OR,\,3.04; 95\% CI, 1.51-6.12) were significantly associated with an increased odds of undergoing imaging in multivariable regression models. CONCLUSIONS: Our observed emergency department imaging rate was 26.5 times higher than our positive result rate, suggesting there is room to decrease unnecessary neuroimaging. Associations for abnormal examination and new headache type are consistent with the American Academy of Neurology clinical imaging recommendations. The increased odds of imaging white patients suggests bias that should be addressed. The low rate of positive findings supports the need for an evidence-based clinical decision tool for neuroimaging in the acute care setting.},
  journal = {The Journal of Pediatrics},
  keywords = {Adolescent,Child,children,Decision Making,Emergency Service; Hospital,Female,Follow-Up Studies,Head,head computed tomography,head magnetic resonance imaging,Headache,Humans,Magnetic Resonance Imaging,Male,migraine,Neuroimaging,Prospective Studies,Tomography; X-Ray Computed},
  language = {eng},
  pmid = {29908647},
}

@Article{Irwin2018,
  title = {Occipital {{Headaches}} and {{Neuroimaging}} in {{Children}}},
  author = {Samantha L. Irwin and Amy A. Gelfand},
  year = {2018},
  month = {jul},
  volume = {22},
  pages = {59},
  issn = {1534-3081},
  doi = {10.1007/s11916-018-0712-6},
  abstract = {PURPOSE OF REVIEW: The goal of this review is to outline the prevalence and significance of occipital headaches in children and how they relate to neuroimaging findings. We seek to evaluate the concern that occipital headaches in children are indicative of secondary headache pathology by reviewing the yield of neuroimaging in pediatric patients with occipital headache location. RECENT FINDINGS: Occipital headaches are a common presentation of primary headache disorders in children, seen in 7-16\% of children presenting for evaluation of headache and in up to 20\% of children diagnosed with migraine in the emergency department. Review of recent literature confirms that in and of itself, occipital location of headache in a child with recurrent headache and a normal physical examination should not be regarded as worrisome. Headaches with associated signs on neurologic examination should be investigated for a secondary cause, regardless of headache location. Occipital headaches that do not meet criteria for a primary headache disorder should be evaluated for site-specific occipital headache conditions. Neuroimaging for recurrent headache in children who have normal neurological examinations has an overall low yield (0-4.1\%) for actionable findings in recent studies. Importantly, an abnormal neurologic examination often predicts the presence of neuroimaging abnormalities. In the absence of an atypical history or abnormalities on clinical examination, occipital headaches in children are no more likely to be associated with intracranial pathology than headaches in other locations. If the child's headaches are otherwise consistent with migraine or another primary headache disorder, and the neurologic examination is normal, the yield of neuroimaging is low, and imaging can generally be deferred.},
  journal = {Current Pain and Headache Reports},
  keywords = {Animals,Child,Headache,Headache Disorders,Humans,Incidental findings,Neuroimaging,Neuroimaging findings,Neurologic Examination,Occipital headache,Pediatric headache,Prevalence,Secondary headaches,Site locked headache},
  language = {eng},
  number = {9},
  pmid = {29987497},
}

@Article{Kan2000,
  title = {Headaches in a Pediatric Emergency Department: Etiology, Imaging, and Treatment},
  shorttitle = {Headaches in a Pediatric Emergency Department},
  author = {L. Kan and J. Nagelberg and J. Maytal},
  year = {2000},
  month = {jan},
  volume = {40},
  pages = {25--29},
  issn = {0017-8748},
  doi = {10.1046/j.1526-4610.2000.00004.x},
  abstract = {OBJECTIVE: To assess the spectrum of diagnoses, the use of CT scans of the brain, and pharmacological treatments in patients presenting to a pediatric emergency department with headaches as the chief complaint. METHODS: A 1-year retrospective chart review of all children who presented to the emergency department with a headache as the chief complaint. RESULTS: One hundred thirty patients (0.7\% of all pediatric emergency department visits, mean age = 9.3 years) were included in the study. Primary headaches included 11 migraine (8.5\%) and 2 tension headaches (1.5\%). Most of the secondary nonneurological headaches were associated with viral and respiratory illnesses (n=37, 28.5\%), while the majority of the secondary neurological headaches included 26 posttraumatic (20\%), 15 possible ventriculoperitoneal shunt malfunctions (11.5\%), and 3 cases of aseptic meningitis (2.3\%). The neurological etiology in 9 of these children (6.9\%) was found to be serious (subdural hematoma, epidural hematoma, proven ventriculoperitoneal shunt malfunction, brain abscess, pseudotumor cerebri, and aseptic meningitis). Fifteen patients could not be etiologically classified, either because no specific etiology was found or their discharge diagnoses could not clearly explain the headache. Approximately 10\% (5/53) of the CT scans of the head that were performed showed new abnormal findings including hydrocephalus secondary to ventriculoperitoneal shunt malfunction (2), subdural hematoma (1), epidural hematoma (1), and skull fracture (1). Forty-two patients (32\%) were treated pharmacologically. Thirty-nine of the 42 treated patients (93\%) were given over-the-counter analgesics, and 9 (21\%) were given prescribed analgesics. CONCLUSIONS: The majority of the headaches in the pediatric emergency department were secondary to concurrent illness and minor head trauma, and required no pharmacological treatment or only treatment with minor analgesics. In a small minority of patients, headaches were secondary to serious neurological conditions, which required immediate medical attention. Computed tomography scans showed new abnormalities in a minority of patients and should be reserved for those with neurological diagnoses such as head trauma and ventriculoperitoneal shunt, as well as for those patients with recent onset of headaches with no clear etiological explanation, and for those with high-risk medical conditions, such as hypocoagulabilities. Future prospective studies are needed to assess the efficacy of the various pharmacological treatments in this population.},
  journal = {Headache},
  keywords = {Adolescent,Analgesics,Child,Child; Preschool,Emergency Service; Hospital,Female,Headache,Humans,Infant,Male,New York,Pediatrics,Retrospective Studies,Tomography; X-Ray Computed},
  language = {eng},
  number = {1},
  pmid = {10759899},
}

@Article{Young2018,
  title = {Neuroimaging Utilization and Findings in Headache Outpatients: {{Significance}} of Red and Yellow Flags},
  shorttitle = {Neuroimaging Utilization and Findings in Headache Outpatients},
  author = {Nathan P. Young and Muhamad Y. Elrashidi and Paul M. McKie and Jon O. Ebbert},
  year = {2018},
  month = {oct},
  volume = {38},
  pages = {1841--1848},
  issn = {1468-2982},
  doi = {10.1177/0333102418758282},
  abstract = {Background Neuroimaging for headache commonly exceeds published guideline recommendations and may be overutilized. Methods We conducted a retrospective cross-sectional study of all outpatient community patients at Mayo Clinic Rochester who underwent a neuroimaging study for a headache indication in 2015. We assessed the neuroimaging utilization pattern, clinical application of red flags, and concordance with neuroimaging guidelines. Results We identified 190 outpatients who underwent 304 neuroimaging studies for headache. The median age was 46.5 years (range 18-91 years), 65\% were female, and most reported no prior history of headache (n\,=\,97, 51\%). A minority of patients had prior brain imaging studies (n\,=\,44, 23\%) and neurological consultations for headache (n\,=\,29, 15\%). Few studies were ordered after consultation with a neurologist (n\,=\,14, 7\%). Seventy-seven percent of patients were documented to have a {"}red flag{"} justifying the imaging study. Abnormal neuroimaging findings were found in 3.1\% of patients with warning flags (5/161); carotid dissection (n\,=\,3) and reversible cerebral vasoconstrictive syndrome (n\,=\,2). An estimated 35\% of patients were imaged against guidelines. Conclusions The prevalence of serious causes of headache in a community practice was low despite the presence of a documented red flag symptom. Inadequate understanding or application of red flags may be contributing to recommendations to image patients against current guidelines. Interventions to reduce unnecessary neuroimaging of patients with headache need to be designed and implemented.},
  journal = {Cephalalgia: An International Journal of Headache},
  keywords = {Adolescent,Adult,Aged,Aged; 80 and over,Cross-Sectional Studies,Female,guidelines,Headache,Humans,magnetic resonance imaging,Male,Medical Overuse,Middle Aged,neuroimaging,Neuroimaging,Outpatients,Practice Guidelines as Topic,Red flags,Retrospective Studies,utilization,Young Adult},
  language = {eng},
  number = {12},
  pmid = {29433347},
}
@Article{Brenner2001,
  title = {Estimated Risks of Radiation-Induced Fatal Cancer from Pediatric {{CT}}},
  author = {D. Brenner and C. Elliston and E. Hall and W. Berdon},
  year = {2001},
  month = {feb},
  volume = {176},
  pages = {289--296},
  issn = {0361-803X},
  doi = {10.2214/ajr.176.2.1760289},
  abstract = {OBJECTIVE: In light of the rapidly increasing frequency of pediatric CT examinations, the purpose of our study was to assess the lifetime cancer mortality risks attributable to radiation from pediatric CT. MATERIALS AND METHODS: Organ doses as a function of age-at-diagnosis were estimated for common CT examinations, and estimated attributable lifetime cancer mortality risks (per unit dose) for different organ sites were applied. Standard models that assume a linear extrapolation of risks from intermediate to low doses were applied. On the basis of current standard practice, the same exposures (milliampere-seconds) were assumed, independent of age. RESULTS: The larger doses and increased lifetime radiation risks in children produce a sharp increase, relative to adults, in estimated risk from CT. Estimated lifetime cancer mortality risks attributable to the radiation exposure from a CT in a 1-year-old are 0.18\% (abdominal) and 0.07\% (head)-an order of magnitude higher than for adults-although those figures still represent a small increase in cancer mortality over the natrual background rate. In the United States, of approximately 600,000 abdominal and head CT examinations annually performed in children under the age of 15 years, a rough estimate is that 500 of these individuals might ultimately die from cancer attributable to the CT radiation. CONCLUSION: The best available risk estimates suggest that pediatric CT will result in significantly increased lifetime radiation risk over adult CT, both because of the increased dose per milliampere-second, and the increased lifetime risk per unit dose. Lower milliampere-second settings can be used for children without significant loss of information. Although the risk-benefit balance is still strongly tilted toward benefit, because the frequency of pediatric CT examinations is rapidly increasing, estimates that quantitative lifetime radiation risks for children undergoing CT are not negligible may stimulate more active reduction of CT exposure settings in pediatric patients.},
  file = {/Users/danielle/Zotero/storage/9ZWAUYN9/Brenner et al. - 2001 - Estimated risks of radiation-induced fatal cancer .pdf},
  journal = {AJR. American journal of roentgenology},
  keywords = {Adolescent,Cause of Death,Child,Child; Preschool,Female,Humans,Infant,Infant; Newborn,Male,Neoplasms; Radiation-Induced,Risk Assessment,Tomography; X-Ray Computed},
  language = {eng},
  number = {2},
  pmid = {11159059},
}

@Article{Feng2010,
  title = {Radiation Dose and Cancer Risk from Pediatric {{CT}} Examinations on 64-Slice {{CT}}: A Phantom Study},
  shorttitle = {Radiation Dose and Cancer Risk from Pediatric {{CT}} Examinations on 64-Slice {{CT}}},
  author = {Shi-Ting Feng and Martin Wai-Ming Law and Bingsheng Huang and Sherry Ng and Zi-Ping Li and Quan-Fei Meng and Pek-Lan Khong},
  year = {2010},
  month = {nov},
  volume = {76},
  pages = {e19-23},
  issn = {1872-7727},
  doi = {10.1016/j.ejrad.2010.03.005},
  abstract = {OBJECTIVE: To measure the radiation dose from CT scans in an anthropomorphic phantom using a 64-slice MDCT, and to estimate the associated cancer risk. MATERIALS AND METHODS: Organ doses were measured with a 5-year-old phantom and thermoluminescent dosimeters. Four protocols; head CT, thorax CT, abdomen CT and pelvis CT were studied. Cancer risks, in the form of lifetime attributable risk (LAR) of cancer incidence, were estimated by linear extrapolation using the organ radiation doses and the LAR data. RESULTS: The effective doses for head, thorax, abdomen and pelvis CT, were 0.7mSv, 3.5mSv, 3.0mSv, 1.3mSv respectively. The organs with the highest dose were; for head CT, salivary gland (22.33mGy); for thorax CT, breast (7.89mGy); for abdomen CT, colon (6.62mGy); for pelvis CT, bladder (4.28mGy). The corresponding LARs for boys and girls were 0.015-0.053\% and 0.034-0.155\% respectively. The organs with highest LARs were; for head CT, thyroid gland (0.003\% for boys, 0.015\% for girls); for thorax CT, lung for boys (0.014\%) and breast for girls (0.069\%); for abdomen CT, colon for boys (0.017\%) and lung for girls (0.016\%); for pelvis CT, bladder for both boys and girls (0.008\%). CONCLUSION: The effective doses from these common pediatric CT examinations ranged from 0.7mSv to 3.5mSv and the associated lifetime cancer risks were found to be up to 0.16\%, with some organs of higher radiosensitivity including breast, thyroid gland, colon and lungs.},
  file = {/Users/danielle/Zotero/storage/DNUFHQHZ/Feng et al. - 2010 - Radiation dose and cancer risk from pediatric CT e.pdf},
  journal = {European Journal of Radiology},
  keywords = {Child; Preschool,China,Female,Humans,Incidence,Male,Neoplasms; Radiation-Induced,Phantoms; Imaging,Radiation Dosage,Risk Assessment,Risk Factors,Tomography; X-Ray Computed,Whole-Body Counting},
  language = {eng},
  number = {2},
  pmid = {20363573},
}
@Misc{UnitedStatesDepartmentofHealthHumanServices:NationalCenterforHealthStatistics2019,
  title = {National {{Hospital Ambulatory Medical Care Survey}}, 2017},
  shorttitle = {{{NHAMCS}} 2017},
  author = {{National Center for Health Statistics}},
  year = {2019},
  month = {nov},
  publisher = {{United States National Center for Health Statistics}},
}
@Misc{UnitedStatesDepartmentofHealthHumanServices-NationalCenterforHealthStatistics2019,
  title = {National {{Hospital Ambulatory Medical Care Survey}}, 2017},
  shorttitle = {{{NHAMCS}} 2017},
  author = {{United States Department of Health \& Human Services- National Center for Health Statistics}},
  year = {2019},
  month = {nov},
  publisher = {{United States National Center for Health Statistics}},
}
@Article{Sheridan2013,
  title = {Diagnostic Testing and Treatment of Pediatric Headache in the Emergency Department},
  author = {David C. Sheridan and Garth D. Meckler and David M. Spiro and Thomas K. Koch and Matthew L. Hansen},
  year = {2013},
  month = {dec},
  volume = {163},
  pages = {1634--1637},
  issn = {1097-6833},
  doi = {10.1016/j.jpeds.2013.07.006},
  abstract = {OBJECTIVE: To describe the variability in diagnostic testing and treatment of headaches in children presenting to the emergency department (ED) with use of a nationally representative sample. STUDY DESIGN: This was a retrospective cohort study using the National Hospital Ambulatory Medical Care Survey during 2005-2009. To assess the use of evidence-based treatment, we analyzed all patients {$<$}18 years old in 2 groups: (1) primary discharge diagnosis of headache and (2) discharge diagnosis of migraine. RESULTS: Four hundred forty-eight sampled ED visits from 2005-2009 represented a national estimate of 1.7 million visits with a discharge diagnosis of headache. A total of 95 visits represented a national estimate of 340\hphantom{,}000 visits with a discharge diagnosis of migraine. Median age was 13.1 years and 60\% were female with a primary diagnosis of headache. In this group, neuroimaging was performed in 37\% of patients and 39\% underwent blood tests. Nonsteroidal anti-inflammatory drugs and opioids were most commonly used for treatment. For children with a discharge diagnosis of migraine, approximately 40\% of patients received non-evidence-based treatment, most commonly with opioid medications, and {$>$}20\% of patients underwent computed tomography scanning. CONCLUSIONS: There is significant variability in the evaluation and treatment of pediatric headache in the ED. Despite evidence-based clinical guidelines for migraine headache, a large number of children continue to receive opioids and ionizing radiation in the ED.},
  journal = {The Journal of Pediatrics},
  keywords = {Adolescent,CBC,Child,Child; Preschool,Cohort Studies,Complete blood cell count,Computed tomography,Cross-Sectional Studies,CT,EBT,ED,Emergency department,Emergency Service; Hospital,Emergency Treatment,Evidence-based treatment,Female,Headache,Humans,Infant,Male,National Hospital Ambulatory Medical Care Survey,NHAMCS,Nonsteroidal anti-inflammatory drug,NSAID,Retrospective Studies},
  language = {eng},
  number = {6},
  pmid = {23968749},
}

@Manual{RStringr,
  title = {stringr: Simple, Consistent Wrappers for Common String Operations},
  author = {Hadley Wickham},
  year = {2019},
  note = {R package version 1.4.0},
  url = {https://CRAN.R-project.org/package=stringr},
}