ABCDEFGHIJKLMNOPQRSTUVWXYZAAABACADAEAFAGAHAIAJAKALAMAN
1
ReferenceLinkLocationInfectionMeasured outcomeInfection CasesDecreasestatistically significant?other important resultsresult adjusted by strenth of evidenceMy impression of relative strength of evidencereasoning behind judgements of strength of evidenceduration of study (years)size intervention groupsize control groupnumber of times the infection was introduced(Class-)room size Ventilation rate (mechanical air changes per hour)Height of fixtureType of lampLamp input or output in WattsIrradiance or dose applied to the upper roomeACH estimates from the upper-room UVGIGood aspects of the study designIssues with the study designOther notes
2
Before/ControlUVGIPercentage
3
Wells et al. 1942 (table 11)https://academic.oup.com/aje/article-abstract/35/1/97/85502?redirectedFrom=fulltextGermantown Friends School, PACold viruses2122173818%0see comment in F34170 m^32,13 mtwo crossed "Safe-t-aire" quartz tubes in a shallow aluminum pan reflector1,34 μW/cm^2 (according to Kowalski 2009 p. 227), original source gives 1,25 mw. per sq. ft>120Besides the classrooms, other shared areas like music room, library, nature room, hallways, lunchroom, rest room and gymnasium were also irradiated; They had access to records of illnesses of the schoolchildren from the 5 years before the studyno real control group; the comparison is between total number of colds over 5 years in un-irradiated primary grades (1-4) rooms from 1933/34 until 1937/38 and total number of colds over 4 years in the same, now irradiated, primary grades (1-4) rooms from 1937/38 until 1940/41, but obviously with other children. Furthermore, the total number of enrolled students in the primary grades changed in those timespansI don’t understand the way they calculated average irradiance intensity: 1.25 mw. per sq. ft. (Milliwatt feet divided by room volume in cubic feet, Milliwatt feet = the summed products of angular flux density and radiant distance in feet)
4
Wells et al. 1942 (table 16)https://academic.oup.com/aje/article-abstract/35/1/97/85502?redirectedFrom=fulltextCombined results for 4 PA schoolsMeaslesSecondary attack rate within class59%24%60%Maybe at the 0,02–0,05 level? (according to MRC 1954 p.40, which references Wells, M. W., 1945)From the timing of the measles cases, they deducted that it spread from the older kids to the younger kids, which is unusual and shows that the disease spread little within the irradiated classrooms of the younger kids.1It seems quite robust because the given percentages are rather conservative: a) if you exclude secondary infections acquired at home (from siblings), the percentage reduction is 76%. b) the results also don’t account for younger children being much more susceptible to measles, having less immunity on average. So the fact that there were fewer infections in the younger kids in irradiated classrooms than in the older kids in unirradiated classrooms is a stronger signal because more susceptible kids mean the disease can usually spread much more easily in younger kids than in older kids. c) From the timing of the measles cases, they deducted that it spread from the older kids to the younger kids, which is unusual and shows that the disease spread little within the irradiated classrooms of the younger kids. d) ad hoc comparisons to how many children were infected in other nearby (un-irradiated) schools showed that those experienced many more cases of measles in their primary departments1301278~195 m^32,13 mGermantown: two crossed "Safe-t-aire" quartz tubes in a shallow aluminum pan reflector; Swarthmore: four 30 watt tubes per classroom. Two reflectors were installed on each end wall at 7 ft. levels, directed the light beam upward at an angle of 60° from horizontal Germantown: 1,34 μW/cm^2 (according to Kowalski 2009 p. 227), original source gives 1,25 mw. per sq. ft>120In Germantown school: besides the classrooms, other shared areas like music room, library, nature room, hallways, lunchroom, rest room and gymnasium were also irradiated; They had access to records of illnesses of the schoolchildren from the 5 years before the study; The attack rate was fairly high due to the 1940-1941 measles epidemicno real control group; the comparison is between younger children in irradiated classes and older children in un-irradiated classes; in Swarthmore schools: Corridors, lunchroom, gymnasium, and other places where children mingle, were not irradiated; the lamps used in the Germantown school and in the Swarthmore schools were quite different As the comparison was made between kids of different ages, it can’t be considered a “real” control group. Yet it is important to note that, on average, younger children are still much more susceptible to measles, having less immunity. So the fact that there were fewer infections in the younger kids in irradiated classrooms than in the older kids in unirradiated classrooms is a stronger signal because more susceptible kids mean the disease can usually spread much more easily in younger kids than in older kids.
5
Sauer et al. 1942TODOThe Cradle, EvanstonRespiratory infection14,54,668%
6
Wells 1943https://ajph.aphapublications.org/doi/abs/10.2105/AJPH.33.12.1436Germantown Friends School, PAMumpsSecondary attack rate within class12283%not available, but probably not significant0,5small sample size189897,00170 m^32,13 mtwo crossed "Safe-t-aire" quartz tubes in a shallow aluminum pan reflector1,34 μW/cm^2 (according to Kowalski 2009 p. 227), original source gives 1,25 mw. per sq. ft>120Other shared areas like music room, library, nature room, hallways, lunchroom, rest room and gymnasium were also irradiated; They had access to records of illnesses of the schoolchildren from the 5 years before the study; The attack rate was fairly high due to the 1940-1941 measles epidemicno real control groupI don’t understand the way they calculated average irradiance intensity: 1.25 mw. per sq. ft. (Milliwatt feet divided by room volume in cubic feet, Milliwatt feet = the summed products of angular flux density and radiant distance in feet)
7
Wells 1943https://ajph.aphapublications.org/doi/abs/10.2105/AJPH.33.12.14363 other PA schools: College Avenue Primary School, College Avenue High School, Rutgers Avenue SchoolTODO: unclear how to interpret the numbers given as evidence in the paper, think about this189-227 m^32,13 mfour 30 watt tubes per classroom. Two reflectors were installed on each end wall at 7 ft. levels, directed the light beam upward at an angle of 60° from horizontal 7,64 milliwatt feet per cubic foot~120no real control group; in Swarthmore schools: corridors, lunchroom, gymnasium, and other places where children mingle, were not irradiatedI don’t understand the way they calculated average irradiance intensity: 1.25 mw. per sq. ft. (Milliwatt feet divided by room volume in cubic feet, Milliwatt feet = the summed products of angular flux density and radiant distance in feet)
8
Schneiter et al. 1944TODONational Training School for Boys, DCRespiratory infection0%
9
Wheeler et al. 1945TODOCamp Sampson Naval Training Station, NYRespiratory infection20%
10
Higgons and Hyde 1947TODOSt. Luke's Hospital, NYRespiratory infection33%
11
Perkins et al. 1947https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1623610/Mexico, Cato-Meridian and Port Byron schools, New York StateMeasles epidemic0%Measles and chickenpox spread more slowly through the irradiated schools than through the controls. No effect on the total incidence of either disease was demonstrated. (according to MRC 1954 p. 39)11-22 μW/cm^2 (according to Kowalski 2009 p. 227)
12
Du Buy et al. 1948TODOTODOTODOTODOTODOTODO
13
Miller et al. 1948TODOGreat Lakes Naval Training Station, ILRespiratory infection19%372 m^2 (according to Kowalski 2009 p. 226)750 W input in the upper room and 840 W input in the lower room (according to Kowalski 2009 p. 226)
14
Willmon, Hollaender and Langmuir 1948TODOTODOTODOTODOTODOTODO
15
Langmuir, Jarrett and Hollaender 1948TODOTODOTODOTODOTODOTODO
16
Bahlke et al. 1949https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1528333/Mexico, Cato-Meridian and Port Byron schools, New York StateChickenpox38,50%28,90%25%Measles and chickenpox spread more slowly through the irradiated schools than through the controls. No effect on the total incidence of either disease was demonstrated. (according to MRC 1954 p. 39)
17
Downes 1950TODOTODOTODOTODOTODOTODO
18
Gelperin, Granoff and Linde 1951TODOTODOminor upper respiratory tractinfectionsTODOTODOTODO93 μW/cm2 (according to Kowalski 2009 p. 227)
19
MRC 1954 (Table 21, p 37)https://babel.hathitrust.org/cgi/pt?id=coo.31924003451584&view=1up&seq=1Southall Elementary schools, EnglandMeaslesSecondary attack rate within class19,25%13,01%32%not available, maybe significant?The effect on the absence rate due to all causes was trivial. It was only about 5 percent lower in total in irradiated schools; Bacterial counts in the air were consistently much lower in irradiated classrooms, showing clearly that the upper-room UVGI had a germicidal effect; “Examination of individual causes of absence suggested that the irradiation probably reduced the number of absences due to certain diseases, by amounts between 15–45 percent. These diseases were mumps and chickenpox, asthma, gastritis and gastro-enteritis, gastric 'flu,’ scarlet fever, acute pharyngitis and tonsillitis, otitis media, and earache (pp. 26); They suspected that the speed of the transmission of measles was slowed down, but that it didn’t affect the total incidence, because measles is so contagious that it ends up infecting everyone susceptible.1The study was rigorously designed and executed and is well documented.; They had a proper control group and a large sample size; The study ran over 3 years; Unfortunately, the average fluence rate applied to the upper room was rather low and declined substantially over the 3 years. Maybe there was no larger effect on disease reduction for that reason.31.5501.860~177 m^3 (20 ft. x 24 ft. with a ceiling height of 13 ft. or over)~6-7 ACH (6,2 ACH in control classrooms and 6,8 ACH in irradiated classrooms)2,4 m or 2,7 m (8 or 9 feet)two tubular 30-watt, low-pressure mercury lamps with their principal output, about 6 watts, at 253.7 nm~2,7 W outputaverage UV intensity in the upper room was 8-13 μW/cm^2; after two years it was only half of that (~4-7 μW/cm^2)not availableSuitable control group (Schools that were similar in almost all relevant aspects, e.g. kids were of the same ages); They made frequent measurements of factors such as ventilation rate, temperature, and humidity and showed that there were no significant differences in these between the intervention and control groups that could be confounders; They made frequent bacteriological investigations of the air in the schoolrooms which consistently showed that airborne bacterial count was much lower in irradiated schools.The average fluence rate in the upper room might have been too low. The average UV intensity in the upper room was measured at 8-13 μW/cm^2, and half of that in later years, whereas NIOSH 2009 recommends 30-50 μW/cm^2; The reason for rates of absence not having been reduced much through the use of upper-room UVGI might have been that the kids were infected in other places apart from school. They were aware of this issue.; Prima facie, the study seems well-designed, rigorous, and well-documented. A huge amount of effort went into it. Unfortunately, it seems plausible that the children were often infected outside of school, so the study didn’t show a big effect. This can’t be confirmed in hindsight, but it is a speculation that many researchers have later voiced about the trial.; It is important to note that the trial was officially planned by a committee of the UK ministry of health, which explains why it was so important and probably received special attention in government circles.
20
MRC 1954 (Table 21, p 37)https://babel.hathitrust.org/cgi/pt?id=coo.31924003451584&view=1up&seq=1Southall Elementary schools, EnglandMumpsSecondary attack rate within class2,49%2,22%11%not available, but probably not significantThe effect on the absence rate due to all causes was trivial. It was only about 5 percent lower in total in irradiated schools; Bacterial counts in the air were consistently much lower in irradiated classrooms, showing clearly that the upper-room UVGI had a germicidal effect; “Examination of individual causes of absence suggested that the irradiation probably reduced the number of absences due to certain diseases, by amounts between 15–45 percent. These diseases were mumps and chickenpox, asthma, gastritis and gastro-enteritis, gastric 'flu,’ scarlet fever, acute pharyngitis and tonsillitis, otitis media, and earache (pp. 26); They suspected that the speed of the transmission of measles was slowed down, but that it didn’t affect the total incidence, because measles is so contagious that it ends up infecting everyone susceptible.1The study was rigorously designed and executed and is well documented.; They had a proper control group and a large sample size; The study ran over 3 years; Unfortunately, the average fluence rate applied to the upper room was rather low and declined substantially over the 3 years. Maybe there was no larger effect on disease reduction for that reason.31.5501.860~177 m^3 (20 ft. x 24 ft. with a ceiling height of 13 ft. or over)~6-7 ACH (6,2 ACH in control classrooms and 6,8 ACH in irradiated classrooms)2,4 m or 2,7 m (8 or 9 feet)two tubular 30-watt, low-pressure mercury lamps with their principal output, about 6 watts, at 253.7 nm~2,7 W outputaverage UV intensity in the upper room was 8-13 μW/cm^2; after two years it was only half of that (~4-7 μW/cm^2)not availableSuitable control group (Schools that were similar in almost all relevant aspects, e.g. kids were of the same ages); They made frequent measurements of factors such as ventilation rate, temperature, and humidity and showed that there were no significant differences in these between the intervention and control groups that could be confounders; They made frequent bacteriological investigations of the air in the schoolrooms which consistently showed that airborne bacterial count was much lower in irradiated schools.The average fluence rate in the upper room might have been too low. The average UV intensity in the upper room was measured at 8-13 μW/cm^2, and half of that in later years, whereas NIOSH 2009 recommends 30-50 μW/cm^2; The reason for rates of absence not having been reduced much through the use of upper-room UVGI might have been that the kids were infected in other places apart from school. They were aware of this issue.; Prima facie, the study seems well-designed, rigorous, and well-documented. A huge amount of effort went into it. Unfortunately, it seems plausible that the children were often infected outside of school, so the study didn’t show a big effect. This can’t be confirmed in hindsight, but it is a speculation that many researchers have later voiced about the trial.; It is important to note that the trial was officially planned by a committee of the UK ministry of health, which explains why it was so important and probably received special attention in government circles.
21
MRC 1954 (Table 21, p 37)https://babel.hathitrust.org/cgi/pt?id=coo.31924003451584&view=1up&seq=1Southall Elementary schools, EnglandChickenpoxSecondary attack rate within class2,71%2,22%18%not available, but probably not significantThe effect on the absence rate due to all causes was trivial. It was only about 5 percent lower in total in irradiated schools; Bacterial counts in the air were consistently much lower in irradiated classrooms, showing clearly that the upper-room UVGI had a germicidal effect; “Examination of individual causes of absence suggested that the irradiation probably reduced the number of absences due to certain diseases, by amounts between 15–45 percent. These diseases were mumps and chickenpox, asthma, gastritis and gastro-enteritis, gastric 'flu,’ scarlet fever, acute pharyngitis and tonsillitis, otitis media, and earache (pp. 26); They suspected that the speed of the transmission of measles was slowed down, but that it didn’t affect the total incidence, because measles is so contagious that it ends up infecting everyone susceptible.1The study was rigorously designed and executed and is well documented.; They had a proper control group and a large sample size; The study ran over 3 years; Unfortunately, the average fluence rate applied to the upper room was rather low and declined substantially over the 3 years. Maybe there was no larger effect on disease reduction for that reason.31.5501.860~177 m^3 (20 ft. x 24 ft. with a ceiling height of 13 ft. or over)~6-7 ACH (6,2 ACH in control classrooms and 6,8 ACH in irradiated classrooms)2,4 m or 2,7 m (8 or 9 feet)two tubular 30-watt, low-pressure mercury lamps with their principal output, about 6 watts, at 253.7 nm~2,7 W outputaverage UV intensity in the upper room was 8-13 μW/cm^2; after two years it was only half of that (~4-7 μW/cm^2)not availableSuitable control group (Schools that were similar in almost all relevant aspects, e.g. kids were of the same ages); They made frequent measurements of factors such as ventilation rate, temperature, and humidity and showed that there were no significant differences in these between the intervention and control groups that could be confounders; They made frequent bacteriological investigations of the air in the schoolrooms which consistently showed that airborne bacterial count was much lower in irradiated schools.The average fluence rate in the upper room might have been too low. The average UV intensity in the upper room was measured at 8-13 μW/cm^2, and half of that in later years, whereas NIOSH 2009 recommends 30-50 μW/cm^2; The reason for rates of absence not having been reduced much through the use of upper-room UVGI might have been that the kids were infected in other places apart from school. They were aware of this issue.; Prima facie, the study seems well-designed, rigorous, and well-documented. A huge amount of effort went into it. Unfortunately, it seems plausible that the children were often infected outside of school, so the study didn’t show a big effect. This can’t be confirmed in hindsight, but it is a speculation that many researchers have later voiced about the trial.; It is important to note that the trial was officially planned by a committee of the UK ministry of health, which explains why it was so important and probably received special attention in government circles.
22
Wells 1955 p. 229https://archive.org/details/airborne-contagionHome for Hebrew Infants, NYVaricella epidemic97%0%100%
23
Wells 1955https://archive.org/details/airborne-contagionMexico and Cato-Meridian schoolsMumps epidemic2355975%
24
Wells 1955https://archive.org/details/airborne-contagionPort Byron SchoolMumps epidemic49%45,90%6%
25
Wells 1955https://archive.org/details/airborne-contagionPleasantville and Mt. KiskoMeasles2272174%
26
Wells 1955https://archive.org/details/airborne-contagionPleasantville and Mt. KiskoChickenpox29710465%
27
McLean 1961TODOLivermore CA veteran’s HospitalInfluenza19%2%89%
28
Kingston et al. 1962TODOTODOTODOTODOTODOTODO
29
Nardell 1988TODOBoston Homeless ShelterTuberculosis78%
30
EPRI 1997TODONorth Central Bronx HospitalTB conversion in staff2,50%1,00%60%
31
National TB Coalition 2001 preliminaryTODONational Homeless SheltersTuberculosis7%
32
Average net decrease41%
33
34
Am J Public Health Nations Health. 1950 Jul; 40(7): 808–812.
doi: 10.2105/ajph.40.7.808
PMCID: PMC1528952
PMID: 15425662
Bacteriologic Studies in Disinfection of Air in Large Rural Central Schools. I. Ultra-violet Irradiation *
F. Wellington Gilcreas and Hazel V. Roberts
35
Bacteriologic Studies in Disinfection of Air in Large Rural Central Schools
II. Ultraviolet Irradiation and Triethylene Glycol Vapor Treatment https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1623071/
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100