Burden of Mortality Attributable to Long-Term Exposure to PM_2.5 in Addis Ababa, Ethiopia: A Health Impact Assessment Using AirQ+

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The proposed study ‘Burden of Mortality Attributable to Long-Term Exposure to PM₂.₅ in Addis Ababa, Ethiopia: A Health Impact Assessment Using AirQ+’ presents an environmental epidemiology assessment of mortality in relation to air pollution, specifically PM₂.₅, in a sub-Saharan territory. The study is methodologically sound, relying on a standardized and internationally adopted modelling tool, and presents significant results, highlighting that air quality is an extremely pressing concern for public health.

 

There are some issues to be addressed to improve the clarity of the manuscript.

 

MAJOR

 

1) Throughout the text, PM₂.₅ concentration values are referred to as ‘exposure’ levels. However, the concept of exposure, in line with IPCC standards, implies the interaction between the hazardous agent and the human population, which is not modelled in the presented study. I therefore suggest revising the term ‘exposure’ and replacing it with a more precise term, such as hazard or pollution concentration.

2) Abstract: attributable deaths are also presented as percentages, but it is unclear whether this refers to attributable deaths over the total mortality count or to a different denominator.

3) In the introduction, the main research gap identified is the poor representativeness of previously existing air quality monitoring networks. The study relies on low-cost sensors, but, as shown in Figure 1, the limited number and uneven spatial distribution of recording sites, with some sensors being extremely close to each other and large parts of the city, particularly in the southern district, left uncovered, is arguably insufficient to fully address this issue. At the same time, a fine-resolution spatial assessment of pollution concentration is not strictly necessary in this study, where no spatial analysis is performed and deaths are cumulatively counted over the territory. In light of the above, I suggest substantially revising the problem statement in the introduction and quantitatively discussing to what extent the proposed recording network outperforms previous solutions.

4) When presenting the distribution of PM₂.₅ concentrations, Table 2 and related text, the authors report mean and standard deviation values. The use of these parameters requires assessment of the normality of the data distribution, which is not presented.

5) Parts of the results text are dedicated to the critical interpretation of recorded data and should therefore be moved to the discussion section, for example lines 233 to 243, 282 to 284, and 300 to 303.

6)The authors refer to long-term effects while considering exposure-response values on a yearly basis. Considering that this implies analyzing deaths occurring within one year of exposure, I suggest revising the temporal reference to ‘medium-term’ effects, to better distinguish these effects from health outcomes that develop after decades of exposure, such as chronic conditions and cancer, which are more consistent with the definition of ‘long-term’.

7) The reported ratios of attributable deaths are extremely high. While the authors report alignment with other territories in a similar geographical context, these ratios are significantly higher than those reported in similar studies in other areas of the world. This requires further elaboration and a deeper comparison with the existing literature, attempting to provide a hypothesis for the reasons behind such high values, in addition to the very reasonable statement that the applied model was developed in different geographical contexts.

8) The presented insights about health effects below recommended thresholds represent a significant point and deserve deeper elaboration, possibly through comparison with the existing literature.

 

MINOR

 

9) Abstract: please define the acronym BAM.

10) I suggest removing Fahrenheit measurements for temperature, as this is not an internationally recognized standard.

11) Please carefully check for typos and leftover draft text, such as line 361.

Author Response

See attached

Author Response File: Author Response.docx

Reviewer 2 Report

Comments and Suggestions for Authors

I really don't have much to say about this paper. It is a reasonably good paper that provides a nice descriptive results of air pollution in Addis using monitors. The authors attempt to provide an example of application predicting excess death associated with pollution exposure. The figures and tables are all relevant and of sufficient quality. 

I would wonder whether the uncorrected PurpleAir data belongs here, but its inclusion does not reduce the quality of the paper. 

Table numbering is inconsistent. For example, the 2023 health impact table is labeled as Table 1, despite appearing after Tables 2 and 3. All tables should be renumbered sequentially.

Figure numbering is inconsistent in the annex. Multiple figures appear to be labeled Figure 7. The authors should check all figure numbers and captions.

The author contribution section contains duplicated text stating that all authors have read and agreed to the published version of the manuscript. This should be corrected.

Several web addresses and data sources contain typographical errors, including references to PurpleAir and AirNow. These should be checked carefully.

The manuscript would benefit from a careful English-language edit. Examples include phrases such as “The administratively of city divided,” “allowed for enhanced the characterization,” “Sites adjacent hospital,” and “consistent with pervious study.”

This stands out however:

The manuscript contains unresolved placeholders and drafting notes, including “please write the name of the author,” and template language. 

""The mean concentration observed in this study 360 (26.76 μg/m³) is lower than the 42.4 μg/m³ reported by (please write the name of the author)"

One might say this strongly suggests that the paper was written with ChatGPT.

Author Response

See attached

Author Response File: Author Response.docx

Reviewer 3 Report

Comments and Suggestions for Authors

This study examined PM₂.₅ concentration patterns and their associated health impacts in Addis Ababa from January 2022 to December 2023. Hourly PM₂.₅ data were collected from nine monitoring stations, comprising seven low-cost PurpleAir sensors and two reference-grade Beta Attenuation Monitors.

Figure 1 shows the Description of the study area: a Sub city of Addis Ababa, and PM2.5 monitoring Stations. Please improve the resolution of the map for better clarity. Also, can the authors explain if these stations are background or roadside stations?

Figure 3 shows the Average diurnal pattern of PM₂.₅ concentration across all monitoring stations, aggregated by day of the week. Can the authors explain why the PM2.5 levels peaked at 7:00 am in the morning?

Table 1 shows the Locations and Land-Use Characteristics of the Selected Air Quality Monitoring Sites in Addis Ababa. However, it is unclear that which one are the EPA reference machine and which is the low cost sensors.

Can the authors also show the QA/QC processes of the low-cost sensors? Were the sensors correctly calibrated with EPA reference methods before being deployed to the sites for monitoring?

Table 2 shows A statistical Summary of PM2.5 concentrations recorded at 9 monitoring sites. It seemed that S7 has the lowest mean and standard deviation, but also significantly less counts compare to the other sites. Can the authors explain this situation?

S6 has a very high standard deviation, which showed that the data is not robust and lots of uncertainty in the data.

S3 shows 0.00 as the minimum, which is technically impossible and must be an error. Can the authors double check their data?

The accuracy of the low costs sensors is not discussed in the limitation section and could be one of the biggest limitation in this study.

Author Response

See attached

Author Response File: Author Response.docx

Round 2

Reviewer 3 Report

Comments and Suggestions for Authors

The authors' response to the reviewer's concern was fair, and I have no further questions.