This is a follow-up to the first post documenting how the coronavirus has differentially impacted communities in Massachusetts, using the state’s weekly public health report for data. That post https://www.masscoronavirus.net/community-spread-in-massachusetts-september-26-2020/ described some of the issues with the weekly report, as well as my methodology for calculating how a particular city and town has contributed to the change in either case incident rates or test positivity rates in the state.
| Table 1: Highest Case Incident Rates and Contribution to Case Incidence Increase | |||||||
| Period Ending October 17, 2020 | |||||||
| Highest Case Incident Rates | Largest Contributors to Weekly Case Incidence Increase | ||||||
| City/Town | 14 Day Case Count | Daily Case Rate per 100,000 | City/Town | 14 Day Case Count | Daily Case Rate per 100,000 | % Contribution | |
| Middleton* | 89 | 61.6 | Boston | 1166 | 12.0 | 6.2% | |
| Lawrence | 567 | 45.9 | Lynn | 244 | 17.3 | 5.0% | |
| Chelsea | 150 | 28.4 | Revere | 230 | 27.0 | 4.2% | |
| Revere | 230 | 27.0 | Gloucester | 77 | 19.2 | 3.6% | |
| Everett | 159 | 23.4 | Lawrence | 567 | 45.9 | 3.5% | |
| Kingston | 41 | 21.6 | Brockton | 202 | 14.7 | 2.7% | |
| Webster | 49 | 20.4 | Chicopee | 76 | 9.5 | 2.2% | |
| Marlborough | 121 | 19.9 | Saugus | 71 | 17.8 | 2.1% | |
| Acushnet | 29 | 19.8 | Methuen | 136 | 18.1 | 1.9% | |
| Gloucester | 77 | 19.2 | Wakefield | 59 | 15.6 | 1.8% | |
| Total/State | 8980 | 9.2 | Total/State | 8980 | 9.2 | 33.2% | |
The three left hand columns focus on the communities with the highest per capita case incident rates for the two-week period ending October 17. The state did make one change to the report, adding an asterisk to communities to indicate where some of the cases are institutional and not directly attributable to the local community. A municipality has an asterisk if it has a long-term care facility, a higher-education facility, or a correctional facility that (1) has had more than ten cases in the past fourteen days, and (2) these cases are more than 30% of cases in the community over the past fourteen days. In the case of Middleton, at the top of the list, Middleton House of Corrections has had a large covid outbreak.
The four right hand columns focus on the communities which contributed the most to the increase in the statewide case incidence rate. As I discussed in the prior post, communities on this list tend to be either large in population (e.g., Boston, which is on the list because of its size), or have had a very large change in case incidence (e.g., Gloucester and Saugus, both with populations of abut 30,000). Boston’s case incidence rate increased by about 8%, but Gloucester’s rate almost tripled. The case incidence rate increased statewide from 8.7 cases to 9.2 cases per 100,000 population per day.
The last column in Table 1 shows the percentage contribution of each community to the statewide increase. This column emphasizes that the case increase has been distributed widely across the state, rather than concentrated in just a few communities.
| Table 2: Highest Positivity Rates and Contribution to Positivity Rate Increase | |||||||
| Period Ending October 17, 2020 | |||||||
| Highest Test Positivity Rate | Largest Contributers to Weekly Positivity Rate Increase | ||||||
| City/Town | 14 Day Test Count | Test Positivity Rate (%) | City/Town | 14 Day Test Count | Test Positivity Rate (%) | Relative Impact (%) | |
| Buckland | 21 | 23.8% | Boston | 173,409 | 0.8% | 6.7% | |
| Middleton* | 1,273 | 8.9% | Lawrence | 9,175 | 8.2% | 5.9% | |
| Lawrence | 9,175 | 8.2% | Lynn | 7,399 | 4.3% | 5.6% | |
| Revere | 6,238 | 4.7% | Revere | 6,238 | 4.7% | 4.3% | |
| Berkley | 395 | 4.3% | Gloucester | 2,953 | 2.8% | 4.3% | |
| Everett | 4,580 | 4.3% | Brockton | 8,417 | 3.0% | 3.5% | |
| Lynn | 7,399 | 4.3% | Chicopee | 4,090 | 2.3% | 2.5% | |
| Chesterfield | 74 | 4.1% | Wakefield | 2,816 | 2.6% | 2.4% | |
| Methuen | 4,520 | 4.0% | Methuen | 4,520 | 4.0% | 2.1% | |
| New Bedford | 5,687 | 4.0% | Saugus | 2,694 | 2.9% | 2.0% | |
| State | 885,959 | 1.24% | Total/State | 885,959 | 1.24% | 39.2% | |
The left three columns in Table 2 focus on the communities with the highest test positivity rate. There are some surprising names on the list, in particular Buckland and Chesterfield. This is because they have high positivity rates with very limited testing – Buckland had five positive cases in the two weeks ending October 17, and Chesterfield had three. These are hardly hotspots. However, four of the top ten communities for case incidence rates are also in the top ten for test positivity rates.
The right-hand columns focus on the communities making the largest contribution to the statewide positivity increase from 1.17% to 1.24%. Boston tops this list as well, even with a very small increase in test positivity from 0.74% to 0.8%, because of its large volume of testing. In the two weeks ending October 17, almost 20% of testing in Massachusetts was in Boston. In contrast, Gloucester and Saugus are not large testers, but each had test positivity rate increases of about 1%. The communities driving the increase in case incidence are also driving the increase in test positivity – the top ten communities in Table 1 and Table 2 are the same.
4 replies on “Massachusetts Community Spread Update October 17, 2020”
are we testing enough? or are we not testing enough of the right groups? can you tell if this latest rise in positivity is due to small group gatherings or the causes given in the spring?” finally what metric or metrics from the state would you like to see that might provide a clearer answer on Covid-19 in Massachusetts.
I’m unhappy that the governor has identified one group of people as driving the increases, implying that people are being irresponsible, when the proof of that is weak. If he knows what’s causing the increases, then do something about it. It seems more likely to me that he is picking easy targets. Of course those between 20 and 50 are driving the increase in cases – they make up the majority of active people, and the largest percentage of the state’s population, too. But what does it even mean to say that? There’s what, 3 million residents in that age group?
Something I think is important to point out to people is that Boston, the city, still tracks testing and cases the “old” way – people are added to the total of tested individuals only once, no matter how many times they are tested. When the college students came back, their tests were added once to the tested total, right? That’s good because the state’s “new” way of doing it just throws everything away, at least in my opinion.
The big thing is, as you’ve mentioned before, is that there are terrible flaws in the hospitalization data provided by the state. Only yesterday was I finally able to understand that my own tracking was ridiculously wrong and, worse, misleading. But hospitalizations are such an important sign of the direction we’re going in with COVID.
I think the governor is referring to the per capita case rates for the 20-50 year old population, not the absolute numbers. For the last four weeks, that cohort has higher per capita case rates than either the under 20 group, or over 50 year old group. It may be because they are “out and about” more, and consequently not irresponsible, but I’m guessing that is what he is referring to.
I think the reporting in the weekly public health report uses standardized metrics for reporting test positivity and case incidence across cities and towns in Massachusetts. I;m not sure what you mean by the “old way”. For the weekly public health report, for cases incidence everybody is counted only once, but for tests if you test multiple times each test is included. Perhaps the City of Boston website is different?
I think the best source of hospitalization data is that from the hospitals themselves. I’d use that data in my age reporting as well, but it is not consistent with the state data, nor do hospitals report anything by age.
As a non-epidemiologist, I can’t comment about whether we’re testing enough. Our testing is way up from the early days of the pandemic. However, about 75% of the testing is repeat testers, and half is from educational institutions. This does tend to drive down the test positivity rate because the repeat testers have low positivity rates compared to first-timers. The number of first-timers have stalled out since mid-August, as the number of repeat testers have continued to grow.
There is no way to tell from the data what is driving the case increase, but I wouldn’t be surprised if small group gatherings is at least partially a cause – based on my personal observations only. And again, I’m not an epidemiologist, but true random testing of the population is the only way to determine how extensive covid is in Massachusetts. That’s a nice theoretical solution, but extremely difficult to implement in practice. In terms of metrics, it would be nice if we could get info from the contact tracers about the origin of the cases – but I don’t know if that is available or could be made public.