Jessica El Halabi1, Nathan Palmer1, Michael McDuffie1, Jonathan J Golub2, Kathe Fox1, Isaac Kohane1, Maha R Farhat3. 1. Department of Biomedical Informatics, Harvard Medical School, Boston, MA, USA. 2. Center for Tuberculosis Research, Johns Hopkins University, Baltimore, MD, USA. 3. Department of Biomedical Informatics, Harvard Medical School, Boston, MA, USA; Division of Pulmonary and Critical Care Medicine, Massachusetts General Hospital, Boston, MA, USA. Electronic address: maha_farhat@hms.harvard.edu.
Abstract
BACKGROUND: A high index of suspicion is needed to initiate appropriate testing for tuberculosis due to its protean symptoms, yet health-care providers in low-incidence settings are becoming less familiar with the disease as rates decline. We aimed to estimate delays in tuberculosis diagnosis and treatment at the US national level between 2008 and 2016. METHODS: In this retrospective observational cohort study, we repurposed private insurance claims data provided by Aetna (Connecticut, USA), to measure health-care delays in tuberculosis diagnosis in the USA in 2008-16. Active tuberculosis was determined by diagnosis codes and the filling of anti-tuberculosis treatment prescriptions. Health-care delays were defined as the duration between the first health-care visit for a tuberculosis symptom and the initiation of anti-tuberculosis treatment. We assessed if delays varied over time, and by patient and system variables, using multivariable regression. We estimated household tuberculosis transmission and respiratory complications after treatment initiation. FINDINGS: We confirmed 738 active tuberculosis cases (incidence 1·45 per 100 000 person-years) with a median health-care delay of 24 days (IQR 10-45). Multivariable regression analysis showed that longer delays were associated with older age (8·4% per 10 year increase [95% CI 4·0 to 13·1]; p<0·0086) and non-HIV immunosuppression (19·2% [15·1 to 60·0]; p=0·0432). Presenting with three or more symptoms was associated with a shorter delay (-22·5% [-39·1 to -2·0]; p=0·0415), relative to presenting with one symptom, as did use of chest imaging (-24·9% [-37·9 to -8·9]; p<0·0098), tuberculosis nucleic acid amplification tests (-19·2% [-32·7 to -3·1]; p=0·0241), and care by a tuberculosis specialist provider (-17·2% [-33·1 to -22·3]; p<0·0087). Longer delays were associated with an increased rate of respiratory complications even after controlling for patient characteristics, and an increased rate of secondary tuberculosis among dependents. INTERPRETATION: In the USA, the median health-care delay for privately insured patients with tuberculosis exceeds WHO-recommended levels of 21 days (3 weeks). The results suggest the need for health-care provider education on best practices in tuberculosis diagnosis, including the use of molecular tests and the maintenance of a high index of suspicion for the disease. FUNDING: US National Institutes of Health.
BACKGROUND: A high index of suspicion is needed to initiate appropriate testing for tuberculosis due to its protean symptoms, yet health-care providers in low-incidence settings are becoming less familiar with the disease as rates decline. We aimed to estimate delays in tuberculosis diagnosis and treatment at the US national level between 2008 and 2016. METHODS: In this retrospective observational cohort study, we repurposed private insurance claims data provided by Aetna (Connecticut, USA), to measure health-care delays in tuberculosis diagnosis in the USA in 2008-16. Active tuberculosis was determined by diagnosis codes and the filling of anti-tuberculosis treatment prescriptions. Health-care delays were defined as the duration between the first health-care visit for a tuberculosis symptom and the initiation of anti-tuberculosis treatment. We assessed if delays varied over time, and by patient and system variables, using multivariable regression. We estimated household tuberculosis transmission and respiratory complications after treatment initiation. FINDINGS: We confirmed 738 active tuberculosis cases (incidence 1·45 per 100 000 person-years) with a median health-care delay of 24 days (IQR 10-45). Multivariable regression analysis showed that longer delays were associated with older age (8·4% per 10 year increase [95% CI 4·0 to 13·1]; p<0·0086) and non-HIV immunosuppression (19·2% [15·1 to 60·0]; p=0·0432). Presenting with three or more symptoms was associated with a shorter delay (-22·5% [-39·1 to -2·0]; p=0·0415), relative to presenting with one symptom, as did use of chest imaging (-24·9% [-37·9 to -8·9]; p<0·0098), tuberculosis nucleic acid amplification tests (-19·2% [-32·7 to -3·1]; p=0·0241), and care by a tuberculosis specialist provider (-17·2% [-33·1 to -22·3]; p<0·0087). Longer delays were associated with an increased rate of respiratory complications even after controlling for patient characteristics, and an increased rate of secondary tuberculosis among dependents. INTERPRETATION: In the USA, the median health-care delay for privately insured patients with tuberculosis exceeds WHO-recommended levels of 21 days (3 weeks). The results suggest the need for health-care provider education on best practices in tuberculosis diagnosis, including the use of molecular tests and the maintenance of a high index of suspicion for the disease. FUNDING: US National Institutes of Health.
Authors: Jean Claude S Ngabonziza; Leen Rigouts; Gabriela Torrea; Tom Decroo; Eliane Kamanzi; Pauline Lempens; Aniceth Rucogoza; Yves M Habimana; Lies Laenen; Belamo E Niyigena; Cécile Uwizeye; Bertin Ushizimpumu; Wim Mulders; Emil Ivan; Oren Tzfadia; Claude Mambo Muvunyi; Patrick Migambi; Emmanuel Andre; Jean Baptiste Mazarati; Dissou Affolabi; Alaine N Umubyeyi; Sabin Nsanzimana; Françoise Portaels; Michel Gasana; Bouke C de Jong; Conor J Meehan Journal: J Clin Tuberc Other Mycobact Dis Date: 2022-01-24
Authors: Sachin R Atre; Jayshri D Jagtap; Mujtaba I Faqih; Yogita K Dumbare; Trupti U Sawant; Sunil L Ambike; Jitendra S Bhawalkar; Sandeep K Bharaswadkar; Padmaja K Jogewar; Ramji S Adkekar; Balasaheb P Hodgar; Vaishali Jadhav; Nitin D Mokashi; Jonathan E Golub; Avika Dixit; Maha R Farhat Journal: Am J Respir Crit Care Med Date: 2022-01-15 Impact factor: 21.405