Influenza follicles Gen Med 2011;12:51-60
• インフルエンザ感染症の患者において,
後咽頭壁にイクラのような濾胞性病変を生じる.
uenza follicles(IF)in seasonal A/H1N1 influenza(Panel A)and type B influen
/
• Definitive influenza follicles; evaluation(Panel C)and on day 3(Pane
al findings in the same patient on initial小型で孤発性のリンパ濾胞(1-2mm).
General Medicine vol. 12 no. 2, 2011
follicles(IF)are flattened, have a broader base, and are cloudy.
半球状で, 境界明瞭, 起始部は広く, くびれは無い.(Y/F type II)
and F show follicles in a febrile disease that is not influenza. The follicles are polym
緊満, 光沢を帯びた表面で, イクラ様と例えられる.
aggregated small nodules. Even those that are round or rice grain-shaped and
follicles are usually not redder than the surrounding pharyngeal mucosa.
7
• Probable influenza follicles; Y/F type I様の形で左右対称性.
周囲の口咽頭粘膜よりも発赤が強く, 内容物は透過性が高い
– 通常インフル感染後数日経ってから生じることが多い.
Diagnosis of Seasonal and Novel Influenza from Pharyngeal Follicles
Figure 6. Yamada/Fukutomi(Y/F type)classification of ele-
/ /
vated lesions(in the stomach)shown schematically with descrip-
tions of physical appearance.
Gen Med 2011;12:51-60
8
• Non-specific follicles; 典型例よりも大きく, 不整形.
様々な大きさのリンパ濾胞が咽頭後壁に認められる.
Y/F typeはII~Iが多い.
– 白みを帯びた赤色のことが多く, 辺縁は収縮した様ないびつ.
– インフルエンザ晩期で認められるが,
他の感染症でも認められることがあるため, 非特異的.
Gen Med 2011;12:51-60
9
迅速検査のMeta-analysis Review
Accuracy of Rapid Intern Med. 2012;156:500 –511.
Ann Influenza Diagnostic Tests
• 159 trialsのmeta. 26種類の迅速検査キットを調査
Table 2. Accuracy Estimates From Subgroup Analyses
Characteristic Pooled Sensitivity P Value Pooled Specificity P Value
(95% CI), % (95% CI), %
Population
Children (60 studies) 66.6 (61.6–71.7) Ͻ0.001 98.2 (97.5–99.0) 0.135
Adults (33 studies) 53.9 (47.9–59.8) Reference 98.6 (98.0–98.9) Reference
Virus type
Influenza A (72 studies) 64.6 (59.0–70.1) 0.62 99.1 (98.7–99.4) Ͻ0.001
Influenza B (27 studies) 52.2 (45.0–59.3) 0.050 99.8 (99.7–99.9) Ͻ0.001
Influenza A and B (47 studies) 62.3 (55.2–69.4) Reference 96.1 (94.4–97.8) Reference
Study conducted during the H1N1 pandemic
Yes (41 studies) 56.3 (48.7–63.9) 0.065 98.9 (98.3–99.5) 0.022
No (74 studies) 65.0 (59.7–70.4) Reference 97.5 (96.6–98.5) Reference
Index test*
BinaxNOW (17 studies)† 57.0 (45.9–67.5) 0.028‡ 98.6 (96.9–99.3) 0.057‡
Directigen Flu A (10 studies) 76.7 (63.8–86.0) 0.49‡ 97.2 (92.6–99.0) 0.62‡
Directigen Flu AϩB (30 studies) 57.2 (48.8–65.2) 0.011‡ 99.3 (98.8–99.6) Ͻ0.001‡
QuickVue Influenza (16 studies) 69.0 (58.1–78.2) 0.66‡ 95.8 (91.3–98.0) 0.82‡
QuickVue Influenza AϩB (21 studies) 48.8 (39.0–58.8) Ͻ0.001‡ 98.4 (96.8–99.2) 0.064‡
Reference standard
RT-PCR (67 studies) 53.9 (48.2–59.6) Ͻ0.001 98.8 (98.3–99.3) 0.002
Culture (48 studies) 72.3 (66.8–77.9) Reference 96.7 (95.2–98.3) Reference
Type of specimen
Nasopharyngeal aspirate (15 studies) 66.6 (56.2–77.0) 0.42§ 97.8 (95.6–100) 0.34§
Nasopharyngeal swab (19 studies) 61.6 (52.0–71.3) 0.75§ 99.1 (98.4–99.9) 0.133§
Nasopharyngeal wash (3 studies) 50.7 (25.1–76.3) 0.32§ 98.1 (94.0–100) 0.82§
Nasal swab (10 studies) 65.9 (53.3–78.5) 0.61§ 99.2 (98.2–100) 0.28§
Throat swab (4 studies) 54.9 (32.7–77.1) 0.45§ 90.0 (74.7–100) 0.018§
Reference standard
RT-PCR (67 studies) 53.9 (48.2–59.6) Ͻ0.001 98.8 (98.3–99.3) 0.002
Culture (48 studies) 72.3 (66.8–77.9) Ann Intern (95.2–98.3) Tests
Med. 2012;156:500Reference
Reference of Rapid Influenza Diagnostic
Accuracy 96.7 –511. Review
Type of specimen
Nasopharyngeal aspirate (15 studies) 66.6 (56.2–77.0) 0.42§ 97.8 (95.6–100) 0.34§
Nasopharyngeal swab (19 studies)
Table 2. Accuracy Estimates From Subgroup Analyses (52.0–71.3)
61.6 0.75§ 99.1 (98.4–99.9) 0.133§
Nasopharyngeal wash (3 studies) 50.7 (25.1–76.3) 0.32§ 98.1 (94.0–100) 0.82§
Nasal swab (10 studies) 65.9 (53.3–78.5) 0.61§ 99.2 (98.2–100) 0.28§
Characteristic (4 studies)
Throat swab 54.9 (32.7–77.1)
Pooled Sensitivity 0.45§
P Value 90.0 (74.7–100)
Pooled Specificity P 0.018§
Value
(95% CI), % (95% CI), %
Testing at the point of care
Population
Yes (28 studies) 58.0 (48.8–67.2) 0.28 97.6 (96.1–99.1) 0.30
Children (60 studies) 66.6 (61.6–71.7) Ͻ0.001 98.2 (97.5–99.0) 0.135
No (91 studies) 63.6 (58.8–68.5) Reference 98.4 (97.7–99.0) Reference
Adults (33 studies) 53.9 (47.9–59.8) Reference 98.6 (98.0–98.9) Reference
Study quality
Virus type
Spectrum of disease
Influenza A (72 studies) 64.6 (59.0–70.1) 0.62 99.1 (98.7–99.4) Ͻ0.001
During influenza season (105 studies) 60.6 (56.0–65.2) 0.032 98.2 (97.6–98.9) 0.62
Influenza B (27 studies) 52.2 (45.0–59.3) 0.050 99.8 (99.7–99.9) Ͻ0.001
Outside influenza season (14 studies) 74.2 (63.9–84.4) Reference 97.8 (95.8–99.8) Reference
Influenza A and B (47 studies) 62.3 (55.2–69.4) Reference 96.1 (94.4–97.8) Reference
Patient selection
ILI defined (45 studies) 59.4 (52.2–66.6) 0.30 97.9 (96.9–99.0) 0.50
Study conducted during the H1N1 pandemic
ILI not defined (74 studies) 64.1 (58.7–69.5) Reference 98.3 (97.7–99.0) Reference
Yes (41 studies) 56.3 (48.7–63.9) 0.065 98.9 (98.3–99.5) 0.022
Blinding
No (74 studies) 65.0 (59.7–70.4) Reference 97.5 (96.6–98.5) Reference
Any blinding reported (54 studies) 61.7 (55.2–68.2) 0.78 97.8 (96.7–98.8) 0.20
No blinding reported (65 studies) 62.9 (57.0–68.7) Reference 98.5 (97.8–99.2) Reference
Index test*
Handling of indeterminate results
BinaxNOW (17 studies)† 57.0 (45.9–67.5) 0.028‡ 98.6 (96.9–99.3) 0.057‡
Reported (19 studies) 66.9 (56.5–77.3) 0.37 98.0 (96.5–99.6) 0.82
Directigen Flu A (10 studies) 76.7 (63.8–86.0) 0.49‡ 97.2 (92.6–99.0) 0.62‡
Not reported (100 studies) 61.5 (56.7–66.2) Reference 98.2 (97.6–98.9) Reference
Directigen Flu AϩB (30 studies) 57.2 (48.8–65.2) 0.011‡ 99.3 (98.8–99.6) Ͻ0.001‡
Industry sponsoring
QuickVue Influenza (16 studies) 69.0 (58.1–78.2) 0.66‡ 95.8 (91.3–98.0) 0.82‡
Sponsored (23 studies) 73.3 (65.3–81.3) 0.007 97.4 (95.5–99.2) 0.24
QuickVue Influenza AϩB (21 studies) 48.8 (39.0–58.8) Ͻ0.001‡ 98.4 (96.8–99.2) 0.064‡
Not sponsored (96 studies) 59.4 (54.6–64.2) Reference 98.4 (97.8–99.0) Reference
Reference standard
ILI ϭ influenza-like illness; RT-PCR ϭ reverse transcriptase, polymerase chain reaction.
RT-PCR (67 studies) 53.9 (48.2–59.6) Ͻ0.001 98.8 (98.3–99.3) 0.002
* See footnote in Table 1 for names of manufacturers of rapid influenza diagnostic tests.
Culture (48Flu A and B and BinaxNOW Influenza A&B were pooled together because statistical tests showed that they
† BinaxNOW studies) 72.3 (66.8–77.9) Reference 96.7 (95.2–98.3) (data not shown).
performed similarly Reference
‡ Reference category is the combination of the other tests.
§Type of specimen is the combination of the other specimens.
Reference category
Article provided a clear definition of the clinical symptoms on 66.6basis of which patients were recruited for the study.
Nasopharyngeal aspirate (15 studies) the (56.2–77.0) 0.42§ 97.8 (95.6–100) 0.34§
Nasopharyngeal swab (19 studies) 61.6 (52.0–71.3) 0.75§ 99.1 (98.4–99.9) 0.133§
Nasopharyngeal wash (3 studies) 50.7 (25.1–76.3) 0.32§ 98.1 (94.0–100) 0.82§
Rapid influenza diagnostic tests performed better
Nasal swab (10 studies) 65.9 (53.3–78.5) had a noticeable effect on99.2 (98.2–100)
0.61§
13
their accuracy. Also, the quality
0.28§
when assessed against viral culture rather than (32.7–77.1)
Throat swab (4 studies) 54.9RT-PCR criteria investigated (patient
0.45§ 90.0selection, blinding, and han-
(74.7–100) 0.018§
(pooled sensitivity, 72.3% [CI, 66.8% to 77.9%] for cul- dling of uninterpretable results) did not have a statistically
Ann Intern Med. 2012;156:500 –511.
Review Accuracy of Rapid Influenza Diagnostic Tests
Table 3. Studies That Provided Data on Effect of Duration of Symptoms on Test Accuracy
Study, Year (Reference) Duration* Sensitivity (95% CI), % Specificity (95% CI), %
Gordon et al, 2009 (69) Day 1 51.9 (40.3–63.3) 98.4 (95.3–99.7)
Day 2 75.1 (68.3–81.1) 97.9 (96.0–99.1)
Day 3 74.2 (62.0–84.2) 97.9 (94.1–99.6)
Day 4 57.9 (33.5–79.7) 98.6 (94.2–100)
Gordon et al, 2010 (68) Ͻ24 h 41.7 (22.1–63.4) 97.9 (88.9–99.9)
Ն24 h 72.1 (59.9–82.3) 98.4 (94.3–99.8)
Keitel et al, 2011 (83)† Յ12 h 35.0 (19.0–55.0) 100 (88.0–100)
12–24 h 66.0 (54.0–76.0) 97.0 (86.0–100)
24–48 h 92.0 (80.0–97.0) 96.0 (82.0–99.0)
Ͼ48 h 59.0 (36.0–78.0) 100 (90.0–100)
Nilsson et al, 2008 (100) 1–3 d 71.4 (58.7–82.1) 100 (95.1–100)
1–5 d 62.8 (51.7–73.0) 100 (96.7–100)
Ͼ5 d 13.8 (3.9–31.7) 100 (90.0–100)
Poehling et al, 2002 (108) Ͻ4 d 100 (63.1–100) 96.6 (90.4–99.3)
Ն4 d 54.5 (23.4–83.3) 98.4 (94.4–99.8)
Stein et al, 2005 (131) Ͻ48 h 58.3 (27.7–84.8) 96.2 (80.4–99.9)
Ͼ48 h 25.0 (12.1–42.2) 98.6 (95.0–99.8)
Stripeli et al, 2010 (132) Ͻ48 h 75.0 (42.8–94.5) 100 (92.1–100)
Ն48 h 65.4 (44.3–84.8) 94.2 (88.4–97.6)
* Duration of clinical symptoms at the time of testing by the rapid influenza diagnostic test.
† Numbers taken directly from the study because there was not enough information to reconstruct the 2 ϫ 2 table.
sensitivity (73.3% [CI, 65.3% to 81.3%]) than studies not 14
Rapid influenza diagnostic tests have a higher sensitiv-
sponsored by industry (59.4% [CI, 54.6% to 64.2%]). ity for detecting influenza A than influenza B. Studies have
Although this difference was statistically significant, sensi- shown that infection with influenza A(H3N2) (the most
• インフルエンザ迅速検査は 1回 2600円
• タミフル 2cp 5日分処方で約3000円
– シーズン(12末~2月)のインフル迅速検査の陽性率は20-40%.
臨床的に疑わしければタミフル処方でも良い.
– 検査すべきは, 仕事の関係, 学童で診断書がいる場合
例えば,
検査前確率 a%とすると,
(2600 x 100) + 3000 x a
=< 3000 x 100
3000a =< 400 x 100
a =< 40/3 = 13.3%
検査前確率 <13.3%ならば,
迅速検査を行う価値あり.
←→ >14%ならば, 迅速せずに
タミフル処方でも全然OK
Ann Intern Med. 2003;139:321-329.
patients were graded as very low quality because of risk of bias or
3 fewer cases (2 to 5 cases) per 1000 patients
imprecision. Table 4 shows the GRADE evidence Med. 2012;156:512-524.
Ann Intern profile,
and Supplement Table 4 (available at www.annals.org)
抗インフルエンザ薬; Meta-analysis
4 fewer cases (1 to 7 fewer cases) per 1000
patients
73 fewer cardiac events (128 fewer to 16 more
shows the results of preplanned subgroup analyses.
A small study in pregnant women (52) reported an
ts events) per 1000 patients OR of 1.27 (CI, 0.07 to 22.16) for death with oseltamivir.
57 fewerOseltamivir, Zanamivir vs Placeboを比較した74 RCTsのMeta
• cardiac events (19 to 80 fewer events)
ts per 1000 patients
Two other small studies (75, 76) were conducted in out-
00 101 fewer events (80 to 126 fewer events) per patient populations with mild uncomplicated influenza but
– Oseltamivir vs No treatmentの比較deaths. The combined results of 5 Japa-
1000 patient-years did not report on
nese studies in patients with confirmed influenza (31, 49,
Outcome Evidence level (n/N; Control vs Oseltamivir) >48hrで内服
55, 74, 75) suggest that inhaled zanamivir may be associ-
mptoms by approximately 23Low(59/242 vs 31/439) with a slightly shorter symptom duration than OR 0.33[0.12-0.86]
死亡リスク hours (CI, 17 ated OR 0.23[0.13-0.43] oral
on the basis of a large SMD (Ϫ0.94 [CI, oseltamivir (7 hours [CI, 2 to 12 hours]; SMD, 0.26 [CI,
Low(1238/100585 0.07 to 0.45]). However, data from another study (76), 0.52[0.33-0.81]
入院率 in outpatients (71), which
66]). One study vs 431/50125) OR 0.75[0.66-0.89] OR
ncluded 肺炎合併
in the meta-analysis,Very showed a which could not be pooled, reported no statistically signif- 0.22[0.15-0.33]
also Low(2111/100449 vs 647/50017) OR 0.83[0.59-1.16] OR
in the duration of illness and a 40% reduc- icant difference in duration of symptoms. The 2 treatments
中耳炎 Low(546/40022 vs 285/38385) OR 0.75[0.64-0.87]
心血管イベント Low OR 0.58[0.31-1.10]
m-effects meta-analysis of oral oseltamivir versus no antiviral therapy based on studies that provided adjusted effect
致命的な副作用 Low RR 0.76[0.70-0.81]
Outcome Patients (Studies), n Pooled Odds Ratio
(95% CI)
Mortality 681 (3) 0.23 (0.13–0.43)
Hospitalization 150 710 (4) 0.75 (0.66–0.89)
Pneumonia 150 466 (3) 0.83 (0.59–1.16)
Otitis media 78 407 (2) 0.75 (0.64–0.87)
Cardiovascular events 100 830 (2) 0.58 (0.31–1.10)
0.0 0.5 1.0 1.5
Favors Oseltamivir Favors No
Antiviral Therapy
Ann Intern Med. 2012;156:512-524.
• Zanamivir vs No treatmentの比較
Outcome Evidence level (n/N; Control vs Zanamivir)
死亡リスク Very Low(5/74 vs 0/13) OR 0.47[0.02-8.97]
入院率 Very Low(69/2411 vs 22/2350) OR 0.66[0.37-1.18]
Very Low(263/2337 vs 301/2337) OR 1.17[0.98-1.39]
Review 肺炎合併 Influenza Treatment: Evidence From Observational Studies
Antivirals for
中耳炎 Very Low(21/2337 vs 25/2337) OR 1.19[0.67-2.14]
• Oseltamivir vs Zanamivir Oseltamivir Versus Inhaled Zanamivir
Table 4. GRADE Evidence Profile for Oral
Outcome Quality Assessment Summary of Findings
Patients Overall Quality Study Event Rates, n/N (%) Relative Effect Anticipated Absolute Effects
(Studies), n* of Evidence (95% CI)
Zanamivir Oseltamivir Risk With Absolute Effect With
Zanamivir Oseltamivir (95% CI)
Mortality 489 (1) Very low†‡§ 0/13 (0) 21/476 (4.4) OR, 1.27 5 deaths per 1000 1 more death (5 fewer
(0.07–22.16)¶ patients** to 95 more deaths)
per 1000 patients
Hospitalization 489 (1) Very low†‡§ 8/13 (61.5) 329/476 (69.1) OR, 1.4 (0.45–4.35) 615 hospitalizations 76 more hospital-
per 1000 patients izations (197 fewer
to 259 more
hospitalizations) per
1000 patients
ICU admissions, mechanical 489 (1) Very low†‡ 3/13 (23.1) 71/476 (14.9) OR, 0.58 231 admissions per 83 fewer admissions
ventilation, or respiratory (0.16–2.18) 1000 patients (185 fewer to 165
failure more admissions)
per 1000 patients
Duration of signs and 1932 (5) Very low due to 760 1172 – – Mean time was 0.26
symptoms risk of bias† SD higher (0.07 to
0.45 SD higher)††
Complications Not measured
Critical adverse events 1045 (1) Very low due to 402 643 Rate ratio, 2.9 7 adverse events 14 more adverse
risk of bias† (0.37–23.05) per 1000 patient- events (5 fewer to
years 165 more events)
per 1000 patient-
years
Viral shedding (persistent 46 (1) Very low due to 4/23 (17.4) 9/23 (39.1) OR, 3.05 174 cases of 217 more cases of
virus) imprecision and (0.78–11.96) 33
persistent virus persistent virus (33
risk of bias§ per 1000 patients fewer to 542 more
cases) per 1000
patients