Abstract INTRODUCTION:

Monitoring coronavirus disease (COVID-19)-related infections and deaths in Brazil is controversial, with increasing pressure to ease social distance measures. However, no evidence of a sustained, widespread fall in cases exists. METHODS We used segmented (joinpoint) regression analysis to describe the behavior of COVID-19 infections in Brazilian capital cities. RESULTS All capitals showed an exponential or a near-exponential increase in cases through May. A decline in reported cases was subsequently noted in 20 cities but was only significant for 8 (29.6%) and was followed in two by a renewed increase. CONCLUSIONS Caution is warranted when considering the relaxation of restrictions.