General characteristics of reviewed studies: We found 13 articles and 10 reports and included eight published articles / preprint manuscripts [16–23] and three reports [13, 24, 25]. From the 13 found articles, three articles were kept for a next version of this review[8–10], one article was excluded because it was not an epidemic estimation study, and one article was excluded because of absent minimum necessary details on methods. From the 10 found reports, three were included, four were excluded because they were not an independent epidemic modeling study, three excluded because of absent minimum necessary details on methods, and one was excluded because it was not an epidemic estimation study. Appendix Table 1 shows the articles and reports found, included, excluded, and exclusion reasoning. Appendix Figure 1 shows the PRISMA studies flow diagram. Four of the eight reviewed articles were published (Moradi et al., Tuite et al., Zareie et al., Zhuang et al.) [18, 20, 21, 23] and the other four were in preprint phase at the time that we reviewed them (Ahmadi et al., Ghaffarzadegan et al., Muniz-Rodriguez et al., Zhan et al.) [16, 17, 19, 22]. All articles were written in English, and all reports were written in Farsi.
We report our findings following the ‘preferred reporting items’ mentioned above.
(1) Epidemic start date and rationale:
Five studies used the presumed official start date of 2020-02-19 (Ahmadi [16], Muniz-Rodriguez [19], Saberi [13], Zareie [21], Zhan [22]). One study (Moradi [18]) reported their estimates starting from 2020-02-20 without mention of the rationale. Ghaffarzadegan reported most of their estimates starting from 2020-01-02, based on unofficial reporting of suspected cases [17]. Haghdoost et al. designated their “Day-zero” as 2020-01-21 [Hijri solar date 1398-11-01], that is 20 days before the presumed official epidemic start date of 2020-02-19 [24]. Many of the predicted outcome values are zero or close to zero in the graphs prior to day 20 of the graphs. However, some of the graphs do seem to show non-zero values for cases or deaths before their day. They maintain that their start date of the epidemic in Iran (2020-01-21) was designated based on “available documentations and epidemiologic analyses”. Mashayekhi et al. did not mention their epidemic start date, and prediction graphs’ time axis showed day zero to 120 or 360 [25]. We made an assumption that their start date was 2020-02-19. Tuite et al. [20] and Zhuang et al. [23] did not use epidemic start date in their estimations.
(2) Epidemic (disease) model type and description:
Most of the studies used sort of compartmental model, however some of them, (Ghaffarzadegan, Zhan, Haghdoost and Mashayekhi) and were more detailed [17, 22, 24, 25]. Three studies used SIR models (Ahmadi [16], Saberi [13], Zareie [21]); three studies used SEIR+ models (Ghaffarzadegan [17], Haghdoost [24], Zhan [22]), and (Mashayekhi [25]) used SLIR+ models, where S denotes Susceptible, E is for Exposed, I is for Infected, R for Recovered, L for Latent, and in any model with a + sign, there are other components for augmentation of model. Four studies did not use explicit disease models (Moradi [18], Muniz-Rodriguez [19], Tuite [20], Zhuang [23]).
(3) Statistical model type, description, and equation(s):
Some of the studies did not mentioned enough details about their statistical methods and did not clearly differentiate between the disease model and the statistical model. Ghaffarzadegan at al. [17], Haghdoost et al. [24], and Mashayekhi et al. [25] used dynamic models. Few studies provided formal representation (equation) of the model.
(4) Model assumptions and their verification:
None of the reviewed studies did explicitly mention all the assumptions, their verification methods, and results of the verification. Most studies did report some details about their assumptions.
(5) Model scenarios’ detailed description:
There were 45 total study-scenario/models for the 11 included studies. Scenario/models per study ranged from 1 to 12 (median 5). All but one of the studies formulated their scenarios based on planned interventions or natural phenomena that could affect disease transmission (seasonality conditions in Ghaffarzadegan [17]). Ahmadi started with statistical models and reasoned backwards about what intervention scenarios could match each statistical model [16].
Ghaffarzadegan had two policy effect scenarios with different levels of efforts to decrease contact rate as well as three seasonality condition options, that amounted to six total scenarios [17]. Haghdoost had four final scenarios, each with levels of isolation for the infected and suspected patients, as they maintained that “to postpone the heavy wave of the disease, the most effective tool is isolation of patients, in a way that the infected and suspected patients would have the least contact with healthy people”. In the early stages of model building, they modeled “the effects of people’s behaviour change and seasonality on disease transmission”, to show the basic or worst model. Then three intervention scenarios with different levels of isolations were added. The people’s behaviour change and seasonality scenarios end only in the basic or worst scenario with no intervention [24]. Mashayekhi has three scenarios, each with different levels of social [physical] contacts and observation of sanitation cautions. As such, Mashayekhi was the only study that considered two modalities of non-pharmacologic interventions [25]. Details of studies’ scenarios are presented in the Appendix.
The following items were not incorporated in any of the scenarios of the included studies: potential vaccine(s), potential pharmacological treatments, changes in case definition, changes in cause of death definition, possibility of reinfection, possibility of mutations or any change in virulence, prevalence of comorbidities and age-stratification of mortality. There were other factors which were considered in some of the studies, such as: testing availability and number of tests performed (Ghaffarzadegan [17]), interventions on social distancing / quarantine (Ghaffarzadegan [17], Haghdoost [24], Mashayekhi [25]), asymptomatic cases (Ghaffarzadegan [17], Mashayekhi [25]), seasonality (Ghaffarzadegan [17], Haghdoost [24]), completeness of reporting cases and deaths to MOHME (Ghaffarzadegan [17], Saberi [13]) and delays in reporting cases and deaths to MOHME (Ghaffarzadegan [17]).
(6) Validation process and findings:
Only one study fully reported their validation process and findings (Ghaffarzadegan [17]): out of sample prediction test RMSPE (Root mean square percentage error) and RMSE (Root mean square error, as well as sensitivity analysis. The model performed better in replicating cumulative cases of infection and death than recovered. Zhan [22] assumed that once the epidemic spreading profile of a given city in Iran was “related” to that of a given city in the historical archive, that observation permitted by virtue of the validity of another study to formulate their optimization to predict the epidemic progression in any given city in Iran. Zhuang [23] mentioned that their estimates with consistent with similar estimates by Tuite [20].
(7) List and sources of model parameters and input data:
List and sources of model parameters are available in the Supplementary electronic material (“Studies’ Methods” tab). In Haghdoost’s study, for number of deaths and cases to start with, assumptions were made that on day-zero, there had been 1080 persons exposed to the virus In Iran (including 75 in Tehran), from which 90 persons had become infected in Iran (including 5 in Tehran) [24]. Four studies (Tuite, Zhuang Haghdoost and Mashayekhi,) [20, 23–25] did not report using number of confirmed cases or confirmed deaths as model input. Among other studies, Ghaffarzadegan used other sources of data, including unofficial reports for number of cases and death and number of performed tests [17].
(8) Model outputs preferably with uncertainty intervals for scenarios:
No study predicted all four outcomes. The most frequently predicted outcomes ranked as cumulative cases (7 studies), daily cases (6 studies), cumulative deaths (4 studies), and daily deaths (1 study). One study predicted two outcomes (cumulative cases and cumulative deaths). Three studies predicted three of the four outcomes. Nine studies provided time-series estimates for number of infected cases and six studies for the number of deaths; two studies (Tuite [20] and Zhuang [23]) reported estimates of cumulative cases for a single point in time.
Forms of outcomes: The intended outcomes and the terminology used in the included studies for the same outcomes, varied across the studies. For daily cases, two distinct groups could be recognized: daily incident cases, and daily prevalent cases. Our designation of daily incident cases included “new cases” reported daily by MOHME, “new cases” predicted by Haghdoost [24], and “daily cases” by Zareie [21]. Our designation of daily prevalent cases included “current cases” (Ghaffarzadegan [17]), “maximum number of cases per day” (Haghdoost [24]), “daily cases” by Mashayekhi [25], and “daily active cases” by Saberi [13]. Active cases are the difference between total cumulative cases with cumulative number of deceased and recovered cases.
Time period of coverage for estimations are very different among the studies; Haghdoost and Mashayekhi covered the longest time periods [24, 25]. Tuite and Zhuang studies are merely based on number of cases originated in Iran and detected in other countries, and each of them provides only one estimate for the number of cases, not across the time [20, 23]. Saberi updated (and updates) the model outputs in a weekly basis; in each round of running model, the most recent data were used to update previous estimates [13].
Table 1 summarizes the findings regarding the methodology used in the reviewed studies. Table 2 shows the estimates of cumulative deaths. Table 3 summarizes the outcomes at the end of month two (2020-04-19) and month four (2020-06-20) after the official epidemic start date. Estimates of cumulative cases, daily deaths and daily cases are demonstrated in Appendix Tables 2, 3, and 4 respectively. Appendix Table 5 demonstrates predictions of peak dates and values of outcomes, and Appendix Table 6 shows predictions of epidemic control dates and values of outcomes.
Figures 1 to 5 demonstrate the reported and estimated outcomes in median scenarios. Figures 1 and 2 show the cumulative deaths and cumulative cases respectively. Figure 3 shows the daily deaths. Figures 4 and 5 show the estimated daily prevalent cases, with and without the estimate form Saberi [13]. That estimate by Saberi, even in the median scenario, had high values compared to other studies. Appendix Figure 2 demonstrates the officially reported cumulative confirmed cases, deaths, and recovered cases, and Appendix Figure 3 shows the daily equivalents. To visualize the quantitative diversity of the studies’ results, we also graphed the reported and worst-scenario estimated cumulative deaths in Appendix Figures 4 and 5, with and without the estimate form Mashayekhi [25]. That estimate by Mashayekhi, was the most extreme prediction among all the studies.
Table 1. Reported items of methodology of the reviewed studies
|
Ahmadi [16]
|
Ghaffarzadegan [17]
|
Haghdoost [24]
|
Moradi [18]
|
Muniz-Rodriguez [19]
|
Mashayekhi [25]
|
Saberi [13]
|
Tuite [20]
|
Zareie [21]
|
Zhan [22]
|
Zhuang [23]
|
Situation of study
|
Published paper
|
medRxiv preprint
|
Full report (Farsi)
|
Published paper
|
medRxiv preprint
|
Summary report (Farsi)
|
Full online report
|
Published paper
|
Preprint version
|
medRxiv preprint
|
medRxiv preprint
|
Epidemic start date
|
20-02-19
|
20-01-02
|
20-01-21
|
20-02-20
|
20-02-19
|
20-02-19 [?]
|
20-02-19
|
N/M (a)
|
20-02-19
|
20-02-19
|
N/M (a)
|
Inputs: Population
|
N/M (a)
|
Yes
|
Yes
|
No
|
N/M (a)
|
Yes
|
N/M (a)
|
N/M (a)
|
N/M (a)
|
N/M (a)
|
Yes
|
Inputs: Cases
|
Yes
|
Yes
|
No
|
No
|
Yes
|
No
|
Yes
|
No
|
Yes
|
Yes
|
No
|
Inputs: Cases (source)
|
MOHME official reports
|
MOHME official reports; unofficial reports
|
NA (b)
|
NA (b)
|
MOHME official reports
|
NA (b)
|
MOHME official reports, WHO, Worldometers
|
NA (b)
|
MOHME official reports
|
MOHME official reports
|
NA (b)
|
Inputs: Deaths
|
Yes
|
Yes
|
No
|
Yes
|
No
|
No
|
Yes
|
No
|
No
|
Yes
|
No
|
Inputs: Deaths (source)
|
MOHME official reports
|
MOHME official reports; unofficial reports
|
NA (b)
|
MOHME official reports
|
NA (b)
|
NA (b)
|
MOHME official reports, WHO, Worldometers
|
NA (b)
|
NA (b)
|
Mazandaran province deaths (20-02-19 to 20-03-06)
|
NA (b)
|
Other input data
|
Number of cured [recovered] cases
|
Number of tests; detected infected travelers and travel data
|
|
|
|
|
|
Exported cases from Iran to other countries; Travel data
|
|
COVID-19 spreading profiles of 367 cities in China
|
Exported cases from Iran to other countries; Travel data
|
Start day of output
|
20-02-19
|
19-12-31
|
20-01-21
|
20-02-20
|
20-02-19
|
N/M (a)
|
20-02-19
|
20-01-01
|
20-02-19
|
N/M (a)
|
20-02-01
|
End day of output
|
20-04-03
|
20-06-30
|
20-05-20
|
20-03-26
|
20-02-29
|
N/M (a)
|
21-02-02
|
N/M (a)
|
20-04-15
|
N/M (a)
|
20-02-24
|
Output length (days)
|
45
|
183
|
121
|
36
|
11
|
360
|
350
|
N/A (b)
|
57
|
N/A (b)
|
24
|
Place
|
Iran
|
Iran
|
Iran and Tehran capital city
|
Iran
|
Iran and 2 multi-province regions
|
Iran
|
Iran
|
Iran
|
Iran
|
Iran and some of the provinces
|
Iran
|
Compartmental model (c)
|
SIR (c)
|
SEIR+ (c)
|
SEIR+ (c)
|
No
|
No
|
SLIR+ (c)
|
SIR (c)
|
No
|
SIR (c)
|
SEIR+ (c)
|
No
|
Statistical method: name
|
Gompertz Differential Equation, VBDGE (d), Cubic polynomial least squared errors
|
Dynamic simulation model
|
Dynamic model
|
Calculating number of cases based on different assumptions for case fatality rate (CFR)
|
Generalized growth mode; Based on the calculation of the epidemic doubling times
|
Dynamic model
|
Classical SIR (C) mathematical model in epidemiology with homogenous mixing assumption
|
N/M (a) (Fraser 2009 study was cited for methods)
|
3-steps model based on the SIR model
|
A data-driven prediction algorithm to find the most resembling growth curve from the historical profiles in China
|
Binomial distributed likelihood framework
|
R0 estimation results
|
1.75
|
2.72 (before starting the interventions)
|
7.24 (at the beginning), 2.58 (after interventions), 1.82 (conditional to isolation of 50% within 3 days)
|
Not used
|
Two methods: 3.6 and 3.58
|
Not used
|
2.37 (for the last 7 days before 20-03-21)
|
Not used
|
Not used
|
Not used
|
Not used
|
Scenarios: number
|
3 (e)
|
6 (f)
|
4 (g)
|
4 (h)
|
2 (i)
|
3 (j)
|
12 (k)
|
6 (l)
|
1
|
1
|
5 (m)
|
Other factors
|
No
|
Yes (n)
|
Yes (o)
|
No
|
No
|
Yes (p)
|
Yes (q)
|
No
|
No
|
No
|
No
|
Outputs: Cases, deaths, both
|
Both
|
Both
|
Both
|
Cases
|
Cases
|
Both
|
Both
|
Cases
|
Both
|
Cases
|
Cases
|
Model validation
|
No
|
Yes
|
No
|
No
|
No
|
No
|
No
|
No
|
No
|
Mentioned but not explained
|
Yes
|
- N/M: Not mentioned
- N/A: Not applicable
- S: Susceptible, E: Exposed, I: Infected, R: Recovered, L: Latent. In any model with a + sign, there are other components for augmentation of model
- VBDGE: Von Bertalanffy differential growth equation
- S1: Gompertz Differential Equation, S2: Von Bertalanffy differential growth equation, S3: Cubic Polynomial Least squared errors
- Six scenarios based on combination of 2 factors: Seasonality (S), and Policy interventions (P)
- S0: Basic scenario (no intervention), only 10% isolation. S1: Worst scenario, minimum (25%) isolation. S2: Medium scenario, medium (32%) isolation. S3: Best scenario, maximum (40%) isolation.
- Four scenarios based on different measures of CFR (S1: 0.3%, S2: 0.5%, S3: 1%, and S4: 2%)
- Based on two different methods to estimate R0
- S1: Ideal scenario, serious distancing. People reduce their social [physical] contacts to 20% of regular level, voluntarily or on a forced basis, after number of cases and deaths have increased, plus close observation of sanitation cautions, so that transmission rate reduces by 65%. S2: Medium scenario, not serious distancing. People reduce their social [physical] contacts only to 20% of regular level, voluntarily, after number of cases and deaths have increased, and other settings are like scenario 1. S3: Worst scenario. People reduce their social [physical] contacts only to 50% of regular level, voluntarily, after number of cases and deaths have increased, plus inadequate observation of sanitation cautions, so that transmission rate reduces only by 40% (instead of 55%), and 60% of people do not observe the sanitation cautions.
- 12 final scenarios, a combination of three options for number of cases and deaths to start with, and four options for the susceptible population size
- Based on six sets of international travel destinations
- Based on selected combinations of (1) Effective catchment population, (2) Detection window 10 or 8 days, (3) 90% or 70% load factors
- Other factors included: Testing availability; Number of tests performed; Social distancing/ Quarantine interventions; Asymptomatic cases; Seasonality; Completeness of reporting cases and deaths to MOHME; Delays in reporting cases and deaths to MOHME
- Seasonality; Social distancing / Quarantine interventions
- Asymptomatic cases; Social distancing / Quarantine interventions
- Completeness of reporting cases and deaths to MOHME
Table 2. Predictions of cumulative deaths for the end of months one to six after the official epidemic start date (2020-02-19)
|
|
Date, Gregorian
|
20-03-19
|
20-04-19
|
20-05-20
|
20-06-20
|
20-07-21
|
20-08-21
|
|
|
Date, Hijri
|
98-12-29
|
99-01-31
|
99-02-31
|
99-03-31
|
99-04-31
|
99-05-31
|
Study 1st author
|
Scenario / model
|
Outcome
|
Value
|
Value
|
Value
|
Value
|
Value
|
Value
|
MOHME official [5]
|
N/A (a)
|
Cumulative deaths
|
1284
|
5118
|
··
|
··
|
··
|
··
|
Ahmadi [16]
|
M1 (b)
|
Cumulative deaths
|
1264
|
··
|
··
|
··
|
··
|
··
|
Ahmadi [16]
|
M2 (c)
|
Cumulative deaths
|
1322
|
··
|
··
|
··
|
··
|
··
|
Ahmadi [16]
|
M3 (d)
|
Cumulative deaths
|
1263
|
··
|
··
|
··
|
··
|
··
|
Ghaffarzadegan[17]
|
S1P1 (e)
|
Cumulative deaths
|
15317
|
44078
|
70462
|
95658
|
··
|
··
|
Ghaffarzadegan[17]
|
S1P2 (f)
|
Cumulative deaths
|
15317
|
41702
|
52937
|
66549
|
··
|
··
|
Ghaffarzadegan[17]
|
S2P1 (g)
|
Cumulative deaths
|
15317
|
44078
|
68383
|
85262
|
··
|
··
|
Ghaffarzadegan[17]
|
S2P2 (h)
|
Cumulative deaths
|
15317
|
41702
|
52937
|
60015
|
··
|
··
|
Ghaffarzadegan[17]
|
S3P1 (i)
|
Cumulative deaths
|
15317
|
44078
|
68383
|
80213
|
··
|
··
|
Ghaffarzadegan[17]
|
S3P2 (j)
|
Cumulative deaths
|
15317
|
41702
|
52937
|
57341
|
··
|
··
|
Haghdoost [24]
|
S0 (k)
|
Cumulative deaths
|
··
|
··
|
30700
|
··
|
··
|
··
|
Haghdoost [24]
|
S1 (l)
|
Cumulative deaths
|
3824
|
9107
|
13450
|
··
|
··
|
··
|
Haghdoost [24]
|
S2 (m)
|
Cumulative deaths
|
2796
|
6231
|
8632
|
··
|
··
|
··
|
Haghdoost [24]
|
S3 (n)
|
Cumulative deaths
|
··
|
··
|
6030
|
··
|
··
|
··
|
Mashayekhi [25]
|
S1 (o)
|
Cumulative deaths
|
759
|
10,316
|
11751
|
11857
|
··
|
··
|
Mashayekhi [25]
|
S2 (p)
|
Cumulative deaths
|
1285
|
33349
|
61322
|
77302
|
86931
|
92620
|
Mashayekhi [25]
|
S3 (q)
|
Cumulative deaths
|
11752
|
97445
|
612953
|
1819392
|
3002721
|
3562136
|
- N/A: Not applicable
- Ahmadi [16] M1: Model 1, Gompertz
- Ahmadi [16] M2: Model 2, Von Bertalanffy growth
- Ahmadi [16] M3: Model 3, Cubic Polynomial
- Ghaffarzadegan [17] S1P1: Seasonality conditions 1 (no effect or status quo) and Policy effect 1 (status quo contact rate). Estimates for 2020-03-19, the end of first month after the epidemic start date, are equal across the six scenarios.
- Ghaffarzadegan [17] S1P2: Seasonality conditions 1 (no effect or status quo) and Policy effect 2 (aggressive efforts to decrease contact rate by half of what it would be otherwise)
- Ghaffarzadegan [17] S2P1: Seasonality conditions 2 (moderate effect; infectivity of the virus decreases linearly from April 1st and halves by June 1st, then stays the same for the rest of the simulation) and Policy effect 1 (status quo contact rate)
- Ghaffarzadegan [17] S2P2: Seasonality conditions 2 (moderate effect; infectivity of the virus decreases linearly from April 1st and halves by June 1st, then stays the same for the rest of the simulation) and Policy effect 2 (aggressive efforts to decrease contact rate by half of what it would be otherwise)
- Ghaffarzadegan [17] S3P1: Seasonality conditions 3 (very strong mitigating effect; infectivity of the virus decreases from April 1st to a quarter of its base value by June 1st, then stays the same for the rest of the simulation) and Policy effect 1 (status quo contact rate)
- Ghaffarzadegan [17] S3P2: Seasonality conditions 3 (very strong mitigating effect; infectivity of the virus decreases from April 1st to a quarter of its base value by June 1st, then stays the same for the rest of the simulation) and Policy effect 2 (aggressive efforts to decrease contact rate by half of what it would be otherwise)
- Haghdoost [24] S0: Basic scenario (no intervention), only 10% isolation. Prediction values for the end of months one and two not available
- Haghdoost [24] S1: Worst scenario, minimum (25%) isolation. While, Haghdoost figure 3 and figure 4 titles read "maximum isolation", Figures 7 and 8 titles also read "maximum isolation" and show the lowest numbers. Therefore, figures 3 and 4 that show the highest numbers are probably the "minimum isolation"
- Haghdoost [24] S2: Medium scenario, medium (32%) isolation
- Haghdoost [24] S3: Best scenario, maximum (40%) isolation. Prediction values for the end of months one and two not available
- Mashayekhi [25] S1: Ideal scenario, serious distancing. People reduce their social [physical] contacts to 20% of regular level, voluntarily or on a forced basis, after number of cases and deaths have increased, plus close observation of sanitation cautions, so that transmission rate reduces by 65%
- Mashayekhi [25] S2: Medium scenario, not serious distancing. People reduce their social [physical] contacts only to 20% of regular level, voluntarily, after number of cases and deaths have increased, and other settings are like scenario 1
- Mashayekhi [25] S3: Worst scenario. People reduce their social [physical] contacts only to 50% of regular level, voluntarily, after number of cases and deaths have increased, plus inadequate observation of sanitation cautions, so that transmission rate reduces only by 40% (instead of 55%), and 60% of people do not observe the sanitation cautions.
Table 3. Lowest and highest predictions at the end of month 2 (2020-04-19) and month 4 (2020-06-20) after the official epidemic start date (2020-02-19)
|
End of month 2 (20-04-19)
|
End of month 4 (20-06-20)
|
Outcome
|
Lowest
value
|
Study-Scenario
|
Highest
value
|
Study-Scenario
|
Lowest
value
|
Study-Scenario
|
Highest
value
|
Study-Scenario
|
Daily deaths
|
125
|
Mashayekhi[25] S1 (a)
|
7839
|
Mashayekhi [25] S3 (b)
|
5
|
Mashayekhi[25] S1 (c)
|
44934
|
Mashayekhi[25] S3 (d)
|
Cumulative deaths
|
3762
|
Ahmadi [16] M5 (e)
|
97445
|
Mashayekhi [25] S3 (f)
|
86931
|
Mashayekhi[25] S2 (g)
|
3002721
|
Mashayekhi[25] S3 (h)
|
Incident daily cases
|
6934
|
Haghdoost [24] S3 (i)
|
13460
|
Haghdoost [24] S1 (j)
|
··
|
No study (k)
|
··
|
No study (l)
|
Prevalent daily cases
|
370
|
Mashayekhi[25] S1 (m)
|
1472165
|
Saberi [13] S3P50 (n)
|
655
|
Mashayekhi[25] S1 (o)
|
17479235
|
Saberi [13] S2P80 (p)
|
Cumulative cases
|
60720
|
Ahmadi [16] M4 (q)
|
1489201
|
Ghaffarzadegan [17] S2P1 (r)
|
1602592
|
Ghaffarzadegan [17] S3P2 (s)
|
2917927
|
Ghaffarzadegan [17] S1P1 (t)
|
- Mashayekhi [25] S1: Ideal scenario, serious distancing. People reduce their social [physical] contacts to 20% of regular level, voluntarily or on a forced basis, after number of cases and deaths have increased, plus close observation of sanitation cautions, so that transmission rate reduces by 65%
- Mashayekhi [25] S3: Worst scenario. People reduce their social [physical] contacts only to 50% of regular level, voluntarily, after number of cases and deaths have increased, plus inadequate observation of sanitation cautions, so that transmission rate reduces only by 40% (instead of 55%), and 60% of people do not observe the sanitation cautions
- Mashayekhi [25] S1: Ideal scenario, serious distancing. People reduce their social [physical] contacts to 20% of regular level, voluntarily or on a forced basis, after number of cases and deaths have increased, plus close observation of sanitation cautions, so that transmission rate reduces by 65%
- Mashayekhi [25] S3: Worst scenario. People reduce their social [physical] contacts only to 50% of regular level, voluntarily, after number of cases and deaths have increased, plus inadequate observation of sanitation cautions, so that transmission rate reduces only by 40% (instead of 55%), and 60% of people do not observe the sanitation cautions
- Ahmadi [16] M5: Model 5, Von Bertalanffy growth
- Mashayekhi [25] S3: Worst scenario. People reduce their social [physical] contacts only to 50% of regular level, voluntarily, after number of cases and deaths have increased, plus inadequate observation of sanitation cautions, so that transmission rate reduces only by 40% (instead of 55%), and 60% of people do not observe the sanitation cautions
- Mashayekhi [25] S2: Medium scenario, not serious distancing. People reduce their social [physical] contacts only to 20% of regular level, voluntarily, after number of cases and deaths have increased, and other settings are like scenario 1
- Mashayekhi [25] S3: Worst scenario. People reduce their social [physical] contacts only to 50% of regular level, voluntarily, after number of cases and deaths have increased, plus inadequate observation of sanitation cautions, so that transmission rate reduces only by 40% (instead of 55%), and 60% of people do not observe the sanitation cautions
- Haghdoost [24] S3: Worst scenario. People reduce their social [physical] contacts only to 50% of regular level, voluntarily, after number of cases and deaths have increased, plus inadequate observation of sanitation cautions, so that transmission rate reduces only by 40% (instead of 55%), and 60% of people do not observe the sanitation cautions
- Haghdoost [24] S1: Worst scenario, minimum (25%) isolation
- No study: No study estimated incident daily cases for end of month 4 after the epidemic start (20-06-20)
- No study: No study estimated incident daily cases for end of month 4 after the epidemic start (20-06-20)
- Mashayekhi [25] S1: Ideal scenario, serious distancing. People reduce their social [physical] contacts to 20% of regular level, voluntarily or on a forced basis, after number of cases and deaths have increased, plus close observation of sanitation cautions, so that transmission rate reduces by 65%
- Saberi [13] S3P50: Scenario 3 (Worst scenario, based on official reports with correction factor of 10 (after Russell [42], 2020-03-30)) with 50 million susceptible population
- Mashayekhi [25] S1: Ideal scenario, serious distancing. People reduce their social [physical] contacts to 20% of regular level, voluntarily or on a forced basis, after number of cases and deaths have increased, plus close observation of sanitation cautions, so that transmission rate reduces by 65%
- Saberi [13] S2P80: Scenario 2 (Medium scenario, based on official reports with correction factor of 5 (after Dr. Rick Brennan [27], Director of Emergency Operations, World Health Organization)) with 80 million susceptible population
- Ahmadi [16] M4: Model 4, Gompertz growth
- Ghaffarzadegan [17] S2P1: Seasonality conditions 2 (moderate effect; infectivity of the virus decreases linearly from April 1st and halves by June 1st, then stays the same for the rest of the simulation) and Policy effect 1 (status quo contact rate)
- Ghaffarzadegan [17] S3P2: Seasonality conditions 3 (very strong mitigating effect; infectivity of the virus decreases from April 1st to a quarter of its base value by June 1st, then stays the same for the rest of the simulation) and Policy effect 2 (aggressive efforts to decrease contact rate by half of what it would be otherwise)
- Ghaffarzadegan [17] S1P1: Seasonality conditions 2 (moderate effect; infectivity of the virus decreases linearly from April 1st and halves by June 1st, then stays the same for the rest of the simulation) and Policy effect 1 (status quo contact rate)
MOHME: Official reports of MOHME for cumulative deaths and cases at 2020 05 05 were 6418 and 101970 respectively, with highest peaks with 158 daily deaths and 3186 daily cases [5].
Cumulative deaths: Lowest and highest predicted cumulative deaths for the end of the second months were 3762 and 97445, and at the end of month four were 11857 and 1819392 respectively.
Cumulative cases: Lowest and highest predicted cumulative cases for the end of the second month were 60720 and 1489201, and at the end of month four were 1602592 and 2917927 respectively
Daily deaths: Lowest and highest values of predicted highest peak of daily deaths (and their dates) were 443 (27-03-2020) and 44,934 (2020-06-20). Only Mashayekhi showed daily deaths predictions [25].
Daily cases: Lowest and highest predicted prevalent daily cases for the end of months 2 were 370 and 10125068, and at the end of month 4 were 5020 and 17146193 respectively. For months 1, 2, and 3, the highest number of predicted incident daily cases was in best scenario of Haghdoost at the end of month 2 (13460 new cases) [24], whereas the MOHME reported 1343 new cases for that date. The lowest was their worst scenario at the end of month 2 (2,272 new cases).
The highest number of predicted prevalent daily cases was in Saberi’s scenario S2P80, with about 17.5 million cases (17479235) as of end of month 4 [13]. Sabri’s estimates had the highest values across all studies within each month: S3P50 for month 1 (1472165), S3P80 for month 2 (10125068) and month 3 (17115184), and S1P80 for months 5 and 6 (17146193 and 7887213) [13]. The highest number of predicted prevalent daily cases was in Mashayekhi best scenario, month 2, with 370 cases. The highest number of predicted prevalent daily cases predicted by Mashayekhi was in worst scenario of Mashayekhi with about 3.5 million cases (3506023) as of end of month 4. It is notable how their numbers of symptomatic and asymptomatic cases compare across scenarios and across months. Lowest and highest predicted prevalent daily cases for the end of months two were 370 (Mashayekhi best scenario) and 10125068 (Saberi worst scenario), and at the end of month 4 were 5020 (Mashayekhi second best scenario) and 17146193 (Saberi’s S1P80 scenario) [13, 25].
Peak dates and control dates: Predicted highest peak value (and date) was 44934 (2020-06-20) for daily deaths, 15239 (2020-04-11) for incident daily cases, 17930000 (2020-06-16) for prevalent daily, and 70711000 (2020-06-25) for cumulative cases synchronous with predicted highest peak of prevalent daily cases. In Haghdoost [24] study, the three peak dates where the same in three scenarios. Values of predicted incident daily cases were similar for the first and second peaks, but for the third peak, the value decrease sizably from scenario 1 (15172) to scenario 3 (9223). Mashayekhi [25] and Ghaffarzadegan [17] also predicted more than one peak.
Three studies predicted the epidemic control (or end) dates and outcome’s values. Two studies predicted the potential date for epidemic to be controlled in April; Ahmadi et al. predicted the “end of the epidemic” on 2020-05-13 with 87000 cumulative cases or on 2020-06-01 with 4900 cumulative deaths (using Von Bertalanffy model) or 11000 cumulative deaths (using Gompertz model) [16]. Haghdoost predicted that with their either medium or best scenarios, the epidemic would be well controlled in month 2 of Hijri solar year 1399 (2020-04-20 to 2020-05-20). Their ‘maximum number of infected people in day’ would be 92100 in middle scenario and 9150 in best scenario [24]. Zhan et al. predicted that if the “authorities continue to impose strict control measures, the epidemic will come under control by the end of April and is expected to end before June 2020, and as the quality of treatment improves, more rapid recovery will be expected” [22]. Beyond the correspondent values of the predicted outcomes, no further criteria or definition of epidemic end or control was provided.