Hora Segura in Lima, Peru

Lima, Peru recently enacted a law to prohibit alcohol sales after 11 AM at commercial venues every day and at bars and clubs sale is prohibited Sun-Thu after 12 AM and after 3 AM on Fri, Sat and holidays.  This is an important law and a step in the right direction to help prevent and reduce alcohol-related injuries and mortality.  We can very well expect motor vehicle collisions (including pedestrians hit by vehicles) and interpersonal violence to decrease if this law is enforced.  Any law such as this should be evaluated to demonstrate its effectiveness.  Several months ago the Municipalidad Metropolitana de Lima (MML) had a press event (http://goo.gl/7ryJp, http://goo.gl/454eg, http://goo.gl/Fkvkw and http://goo.gl/D1imi)  lauding this law’s effectiveness based on road traffic incidents being reduced by 52% and fatalities by 40% between two time points before and after the law was enacted; however, the statistics and methods they used to back up their claims are very tenuous.  Using such statistics may do this law more harm than good because it casts doubt on its effectiveness.

There several reasons we should be cautious about these statistics.  First, the data come from an active surveillance system run by MINSA, the Ministry of Health.  It is a rather new system that has a good start, but is not a complete surveillance system since it only covers public hospitals, mostly in Lima, so it would only capture road traffic incident victims that visited those hospitals for medical care.  This would most likely include people with at least moderate to severe injuries, but not all of them would go to a public hospital.  Many would go to a private clinic or hospital, thus this system only captures a fraction of the victims.  The press reports cited report that the system registered a total of 150 MVC crash victims in December of 2011 (before the law was enacted) and 70 in April of 2012, only in the district of the Cercado of Lima (the central area of Lima).  These two numbers were used to calculate the reduction of 52% and 40%.  These numbers are likely underreporting the total number of victims since the police in 2006 (the most recent year of data available for this district from the police) reported 243 incidents in December and 213 in April.  We know overall the road traffic incidents in Peru are increasing based on temporal trends reported by the police (there were 1543 incidents from Jan-Jun in 2011 and 1481 in the same months in 2006 in the Cercado http://goo.gl/4yDNz), so it is likely the number of victims is still as high or more than what was reported by MINSA’s surveillance system.  We have to remember too that the police do not catch all road traffic incidents.  Some incidents are unreported because the people involved resolve the matter between themselves or if someone is in a single-vehicle collision they may not report it either.

Second, the numbers reported are not specific to the hours of prohibition of Hora Segura during which we might expect to see an effect.  They are the total number of incidents recorded.  We could reasonably expect a reduction in the number of road traffic incidents from 11 PM to 5 AM Mon-Fri and perhaps from 3-7 AM on Sat & Sun (from drinking from the previous evening/early hours of these days).  If we refer again to the 2006 data, there were 20 victims in Dec and 22 in April during the relevant hours – about 9% of the totals.  We cannot honestly apply the reductive effect of Hora Segura to all road traffic indicents in these two months.

Third, we are only observing two snapshots of road traffic incident victims, not the entire temporal trend.  Incidents can vary over time from month-to-month.  You may randomly get one month with seemingly high number of incidents and another with a low number of incidents.  Comparing the two we might think we are observing a trend, but this unfair because we are not observing the entire trend.  We need month-to-month statistics on the number of victims long before and after the passage of this law to account for temporal trends.  It could be the number of victims is decreasing before the law went into effect, that either month was an aberration, or that something else may account for the reduction. Additionally, these trends are not accounting for the number of drivers, pedestrians, passengers and other road users.  Comparing these numbers alone assumes all else is constant, but that is a strong assumption.

The issue is that the MML could have easily obtained fairly detailed records from the police commissaries that cover the Cercado rather than relying on shoddy statistical tactics.  We could have had a much more complete picture.  Now that we are over a year out from Hora Segura, we could get a better picture of its effect if we went into the commissary records to really get much more complete numbers on how 2011 and 2012 looked in terms of not only road traffic incidents during the specific hours of interest, but also other types of alcohol-related problems like assaults, homicides, suicides and such.