In this outbreak, finding the first case cut the median reporting delay from thirty-three days to eight. But even eight days can overlap the entire infectious period, leaving control measures chasing transmission instead of stopping it. Emerging arboviral diseases in Europe pose a challenge because cases can be difficult to detect and diagnose during the initial circulation of the pathogen.

Early outbreak detection enables public health authorities to take effective actions to reduce disease transmission. That makes quantifying reporting delays vital for planning and assessing surveillance and control strategies. This study provides estimates of reporting delays during an emerging arboviral outbreak and indications of how delays may have impacted onward transmission.

Timely detection of arboviral introduction and transmission is paramount for decreasing the probability of further autochthonous transmission in non-endemic countries. When autochthonous transmission occurs despite efforts, early detection of cases helps bring the outbreak under control.

Because local outbreaks transmitted by Aedes albopictus are still infrequent in Europe, little is known about surveillance performance during an ongoing emergency. The study investigates reporting delays during the 2017 Italian chikungunya outbreak and provides estimates useful for modeling outbreak responses.

Timely detection of infected cases enables public health authorities to reduce disease transmission and limit health and socio-economic impacts. For mosquito-borne diseases, the longer an undetected viremic individual remains exposed to mosquito bites, the higher the risk of subsequent transmission and pathogen spread.

The reporting delay is the period between symptom onset and notification to public health authorities, and its distribution can indicate surveillance quality and health-system preparedness. Its length can vary across patients, physicians, testing laboratories, and the socio-technological infrastructure used to report identified cases.

In 2017, epidemiological investigations and active case finding identified a total of 499 probable cases, suggesting three main foci in Lazio: Anzio, Rome, and Latina, plus Guardavalle in Calabria. The Lazio outbreak was identified when national laboratories analyzed serum and urine samples from a cluster of three suspected cases.

The Lazio surveillance system included passive surveillance based on medical practitioners reporting suspected cases, laboratory-based surveillance, and syndromic surveillance. Reporting suspected arbovirus cases to local public health authorities was mandatory, but reporting time had not been analyzed despite its importance for future preventive measures.

The study analyzes reporting delays during the 2017 chikungunya outbreak in the Lazio region of Italy and discusses potential consequences for controlling disease spread. That focus connects the timing of notification with the practical question of how well outbreak control can work.

The analysis included all suspected, probable, and confirmed chikungunya cases notified by SERESMI between September sixth, 2017, and February fifth, 2018. A suspected case required fever and joint pain, no recent travel to an endemic country, or close contact with a probable or confirmed case.

A probable case was a suspected case positive for anti-CHIKV IgM on a single serum sample, while a confirmed case required CHIKV PCR or additional confirmation criteria. A medical practitioner had to report a suspected arboviral infection within twelve hours and collect samples for laboratory confirmation.

Reporting delay was defined as the number of days between the notification date and the date of symptom onset. Each case was classified as before outbreak detection, or BOD, or after outbreak detection, or AOD, by comparing symptom onset with the first notification on September sixth, 2017.

Reporting delays were investigated with Kaplan-Meier curves using the BOD or AOD classification, and the effect of outbreak detection was also analyzed with a Cox proportional hazard model. The analysis excluded undetected cases because information was available only for notified cases, and sensitivity analyses excluded delays longer than thirty days or considered only confirmed cases.

The study also investigated the proportion of cases notified before the end of their infectious period and estimated the days of uncontrolled onward transmission that notification might have avoided. For each case, the infectious period was modeled to begin one or two days before symptom onset, with equal probability, and to last uniformly between five and eight days.

This process was repeated one thousand times for each case to obtain a distribution of infectious periods. A sensitivity analysis instead fitted a truncated Poisson distribution to fever duration in identified imported chikungunya cases in Italy. A total of 414 cases were notified in Lazio: 312 linked to Anzio, 73 to Rome, 8 to Latina, and 21 scattered across the region.

The cases included 200 confirmed, 202 probable, and 12 suspected cases, while two cases were excluded because symptom-onset information was missing. The investigation found that 189 patients, or 45.9 percent, developed symptoms before outbreak detection, while 223, or 54.1 percent, developed symptoms afterward.

The median reporting delay was 15.5 days overall, 33 days before detection, and 8 days after detection. Figure one reconstructs when cases developed symptoms and were notified across the Anzio, Rome, Latina, and other foci. In panel A, each point marks notification, while the horizontal line shows the reporting delay; the grey region marks the period before outbreak detection, with the earliest retrospectively identified onset occurring on 26 June in Anzio.

Panel B converts these delays into cumulative curves, showing how quickly cases accumulated by reporting delay and making the shift in notification timing around outbreak detection visible. Figure two compares the outbreak by symptom-onset week and notification week, with colors separating the Anzio, Rome, Latina, and other epidemiological foci.

The lower panels show reporting delays as medians, interquartile ranges, minimums, maximums, and means, revealing how delays vary across weeks and become more dispersed when organized by notification date. The dashed lines mark outbreak detection, while the paper reports that notification was about three point five times higher after detection among cases with symptom onset after detection than before it.

Kaplan-Meier curves estimated a median reporting delay of 33 days before outbreak detection and 8 days after outbreak detection. The Cox model estimated a notification hazard ratio of 3.45 after detection compared with before detection, with a 95 percent confidence interval from 2.80 to 4.25.

Sensitivity analyses confirmed a shorter reporting delay after outbreak detection, with the lower delay mainly associated with excluding patients whose reporting delay exceeded thirty days. Among 277 records after that exclusion, 196, or 70.8 percent, concerned cases whose symptom onset occurred after outbreak detection.

The median reporting delay was 9 days overall, 18 days before detection, and 7 days after detection. Kaplan-Meier estimates were 18 days before detection and 7 days after detection, confirming reduced time to notification after detection. Using an infectious period beginning one or two days before symptom onset and lasting five to eight days, an average of 17 percent of cases had notification times shorter than their infectious periods.

That percentage was almost zero before outbreak detection: 0.9 percent, compared with 30.7 percent after detection. The estimated reduction in days per person when onward transmission could occur was zero before detection, 0.3 days per person after detection, and 0.28 days per person across the whole period.

The study identifies the delay between symptom onset and notification as a critical parameter in outbreak control and notes that such delays had been suggested as a key driver of the Lazio outbreak's size and spread. Even after outbreak detection, the median delay was 8 days, comparable to the infectious period of 5 to 8 days.

Because case-based interventions can occur only after case detection, these delays significantly impair control measures' ability to disrupt pathogen transmission. Under the assumption that notification prevents further transmission, the percentage reduction in days when onward transmission could occur was still only 8 percent on average.

The study has limitations: all estimates were computed only from notified cases because data on the total infected population were unavailable. That limitation means the analysis does not directly estimate infections that were never notified.

The paper states that autochthonous transmission events of chikungunya and related arboviruses in Europe are likely to increase because more infected travelers are entering European countries and mosquito populations such as Aedes albopictus are expanding. Europe is likely to face more frequent locally transmitted chikungunya and related arbovirus events in the future, driven by infected travelers and expanding mosquito populations.

The central lesson is that outbreak detection sharply improved reporting speed, but not enough to reliably interrupt transmission. Faster early detection and stronger vector control are therefore essential when chikungunya appears in a non-endemic setting.