Diversity in HIV epidemic transitions in India: An application of HIV epidemiological metrices and benchmarks.

BACKGROUND: The Joint United Nations Programme on AIDS (UNAIDS) has emphasized on the incidence-prevalence ratio (IPR) and incidence-mortality ratio (IMR) to measure the progress in HIV epidemic control. In this paper, we describe the status of epidemic control in India and in various states in terms of UNAIDS's recommended metrices.
METHOD: The National AIDS Control Programme (NACP) of India spearheads work on mathematical modelling to estimate HIV burden based on periodically conducted sentinel surveillance for providing guidance to program implementation and policymaking. Using the results of the latest round of HIV Estimations in 2019, IPR and IMR were calculated.
RESULTS: National level IPR was 0.029 [0.022-0.037] in 2019 and ranged from 0.01 to 0.15 in various States and Union Territories (UTs). Corresponding Incidence-Mortality Ratio was at 0.881 [0.754-1.014] nationally and ranged between 0.20 and 12.90 across the States/UTs.
CONCLUSIONS: Based on UNAIDS recommended indicators for HIV epidemic control, namely IPR and IMR; national AIDS response in India appears on track. However, the program success is not uniform and significant heterogeneity as well as expanding epidemic was observed at the level of States or UTs. Reinforcing States/UTs specific and focused HIV prevention, testing and treatment initiatives may help in the attainment of 2030 Sustainable Development Goals of ending AIDS as a public health threat by 2030.

Introduction

The goal of ending the AIDS epidemic by 2030 is integral to attaining the third Sustainable Development Goal (SDG) of ensuring healthy lives and promoting well-being for all at all ages [1]. HIV incidence rate per 1000 uninfected population is the main indicator to measure progress on the HIV/AIDS response in the era of the SDG [2]. While no benchmark for incidence rate has been recommended to be achieved by 2030 for ending the AIDS epidemic, Joint United Nations Programme on HIV/AIDS (UNAIDS) has called to achieve a 90% decline in annual new HIV infections and AIDS-related deaths by 2030 from the baseline value of 2010 [3-5]. Characteristically, between 2010 and 2019, the number of new HIV infections and AIDS-related deaths have declined by 23% and 39% respectively globally [6].

The target of 90% percentage reductions in new HIV infections and AIDS-related deaths as an indicator of progress towards ‘ending AIDS as a public health threat’ by 2030 has limitations [7, 8]. This indicator does not factor in the epidemic heterogeneity across the globe and ignores the fact that attaining such reductions in low-level epidemic settings would be relatively difficult. Moreover, using the 2010 baseline for these indicators pose a relative disadvantage for the countries having mature interventions who already have achieved strong gains before the baseline year of 2010. In contrast, countries that have scaled up their HIV responses after 2010 have a relative advantage. Further, presenting the progress on new infections and AIDS-related deaths does not adequately depict the association between mortality among people living with HIV (PLHIV), new HIV infections, and the prevalence of HIV making the overall progress on the epidemic.

Recognising these limitations, there have been efforts to search for indicators to measure the progress on the 2030 SDG of ‘ending the AIDS epidemic as a public health threat’. Two epidemiological indicators of the sustainability of transmission have been identified as more refined [4, 7–9]. These are: (i) incidence-prevalence ratio (IPR), and (ii) incidence-mortality ratio (IMR). Assuming an average life expectancy of 30 years after a person acquires HIV infection, attainment of a benchmark value of <0.03 on IPR indicates that the PLHIV number will gradually decrease as there are fewer than three new HIV infections per 100 people living with HIV per year and the ending of AIDS epidemic will be achieved. Attainment of benchmark value of <1 for IMR will indicate that the PLHIV number will gradually decrease as there are fewer new infections than deaths. In 2019, UNAIDS reported that 25 countries had achieved the IPR of <0.03 while fewer countries reported achieving desired threshold of IMR [6].

India, with an estimated PLHIV size of 2.35 million and adult prevalence of 0.22% in 2019, is the second-largest HIV/AIDS epidemic in the world [6, 10]. Initiatives of the National AIDS Control Programme (NACP) in the country resulted in a 37% decline in new HIV infections and a 66% decline in AIDS-related deaths between 2010 and 2019. However, no data are available on IPR and IMR from India. In this article, we report the present status of HIV epidemic control in India using the UNAIDS recommended indicators of IPR and IMR at the national and state levels.

Methods

HIV Surveillance and Estimation is an integral part of the spectrum of activities of the National AIDS Control Organization (NACO) of the Government of India under NACP. This activity is periodically conducted using a specially created programmatic framework built on the foundation of government institutes and partners. As part of the surveillance activity, when primary data and sample collection is done from the survey participants, informed consent is taken from them in alignment with national guidelines. The institutions involved in primary data collection routinely submit their proposals for the surveillance program to their respective ethics committees to seek approval and the survey at each site is initiated only after the local ethics committee approves the proposal. Utilizing this program generated data for policy making and programmatic improvement, including the HIV burden estimation, is a mandate of the NACP in India.

HIV burden estimation is carried out after each round of Surveillance by NACO by making use of aggregated de-identified data. This work has been published periodically in the past. It may be noted that primary data collection from human subjects is not part of HIV burden estimation exercise. The current manuscript is based on the analysis of aggregated de-identified outputs generated through the HIV Estimations 2019 model for each of the Indian State or Union Territory.

We used the modelled estimates for the year 1990 to 2019 from HIV Estimations 2019, the latest round under the NACP, to construct national and state measures of progress on IPR and IMR. The periodic HIV burden estimation exercise under NACP was undertaken employing the UNAIDS supported Spectrum Software (Avenir Health, Glastonbury, Connecticut, USA). The details of the process and method for the same had been described elsewhere [10-17].

In brief, the Spectrum mathematically model demographics, treatment coverage and HIV prevalence data to estimate incidence trend which is first distributed by age, sex and CD4+ counts and then the newly infected population are transitioned over time through age, CD4+ count and treatment (or lack of treatment) categories with death as the final outcome. The ‘Uncertainty Analysis’ tool in Spectrum generates plausible range of key HIV indicators by running 1000 Monte Carlo iterations combining uncertainty in adult incidence produced by EPP with uncertainty around other key assumptions such as fertilty, incidence, mortality etc based on global or regional values. The tool exports estimates for each of 1000 iteration for key indicators.

In India, the estimation process has been designed to develop sub-national level (State/ Union Territory) models where State/ Union Territory-specific demographic, treatment coverage, prevalence and surveillance data are inputted. Epidemiological parameters such as patterns of incidence, progression, mortality, and fertility derived from scientific studies have been in-built in the Spectrum Software while computing the desired outputs.

The Spectrum software is updated periodically under the guidance of the UNAIDS Reference Group on Estimates, Modeling and Projections. The details of the updates are available on the website of reference group (www.epidem.org) and the software developer website (https://avenirhealth.org/software-spectrum.php). HIV Estimations 2019 under NACP in India was implemented using the 5.80 version of Spectrum Software.

For the current study, we used aggregated deidentified outputs for all States/ Union Territories of India generated through HIV Estimations 2019. This included year-wise data from 1990–2019 on annual new HIV infections and annual total deaths among people living with HIV (PLHIV). Using this data we calculated IPR as the ratio of new HIV infections over PLHIV in a given year for a given geography and IMR as the ratio of new HIV infections over all-cause mortality among PLHIV in a given year for given geography as per standard definitions [4].

The uncertainty bound for a State/UT for IPR/IMR was estimated using each of the 1000 iteration values of new HIV infections, PLHIV size and total deaths among PLHIV generated through the ‘Uncertainty Analysis’ tool. After calculating the IPR/IMR for each of the iteration for a State/UT, we obtained the 2.5% and 97.5% percentiles of the 1000 ratios to inform the 95% uncertainty bound estimation for the given State/UT.

To inform the uncertainty bound around national IPR/IMR, we summed the given indicator from all State/UTs by each of 1000 iteration to produce 1000 estimates of incidence, prevalence and mortality for the country. Then we calculated IPR/IMR for each of the 1000 aggregated iterations which were used to generate the uncertainty bound for national estimates.

The current study relied on the analysis of aggregated de-identified outputs generated through the HIV Estimations 2019 model of each State/UTs. As there was no primary data collection for the current analysis, the ethics review was not sought.

Results

Overall, 2.38 million people were estimated to be living with HIV (PLHIV) in 2019 in India. States of Andhra Pradesh, Karnataka, Tamil Nadu, Telangana (in southern India), Maharashtra, Gujarat (in western India), Punjab (in northern India), Uttar Pradesh (in central India), Bihar, West Bengal (eastern India) are the 10 top-ranking States in India in terms of PLHIV size (Fig 1, S1 Table).

Fig 1

Annual new HIV infections, annual all-cause mortality among PLHIV and total number of PLHIV by select States in India, 1990–2019.

The years (1990–2019) are reflected on X-axis; number of new infections (in 1000s) and total deaths and PLHIV (in 1000s) are on primary Y-axis while total number of PLHIV (in 1000s) is on secondary Y-axis. Orange line depicts new infections, grey line depicts total deaths among PLHIV, and blue line represents total PLHIV size.

Nationally, the incidence: prevalence ratio was at 0.029 [0.022–0.037] in 2019 while the incidence: mortality ratio was at 0.881 [0.754–1.014] (Table 1, Figs 2 and 3). The IPR had an overall declining trend nationally with estimates of 0.098 [0.076–0.120] in 2000, 0.041 [0.034–0.049] in 2010, and 0.029 [0.022–0.037] in 2019. Nationally, the IMR declined to 0.481 [0.437–0.530] in 2007 and thereafter had a gradual upward trend to 0.569 [0.501–0.668] in 2010 and to 0.881 [0.754–1.014] in 2019. State/ UT-wise IPR and IMR for the period 1990–2019 may be seen at supporting information at S2 Table.

Table 1

Incidence-prevalence ratio and incidence-mortality ratio (with uncertainty bounds) by States/UTs in India, 2019.

State/UT	IPR			IMR
	Point Estimate	Lower Bound	Upper Bound	Point Estimate	Lower Bound	Upper Bound
Andaman & Nicobar Islands	0.041	0.024	0.093	0.952	0.478	1.972
Andhra Pradesh	0.009	0.004	0.017	0.199	0.099	0.338
Arunachal Pradesh	0.100	0.061	0.122	3.476	1.862	4.742
Assam	0.059	0.049	0.073	1.790	1.512	2.313
Bihar	0.057	0.037	0.070	2.358	2.003	2.930
Chandigarh	0.056	0.029	0.074	1.180	0.681	1.546
Chhattisgarh	0.068	0.052	0.079	1.545	1.316	1.712
Dadra & Nagar Haveli	0.108	0.059	0.127	6.727	3.376	9.474
Daman & Diu	0.063	0.041	0.079	2.875	1.816	3.707
Delhi	0.045	0.035	0.057	2.174	1.798	2.741
Goa	0.010	0.003	0.028	0.281	0.089	0.534
Gujarat	0.031	0.025	0.038	1.378	1.161	1.663
Haryana	0.056	0.043	0.067	1.160	0.921	1.376
Himachal Pradesh	0.022	0.018	0.028	1.235	0.929	1.673
Jammu & Kashmir	0.048	0.028	0.091	1.831	1.304	3.672
Jharkhand	0.055	0.040	0.071	2.192	1.783	3.027
Karnataka	0.006	0.003	0.011	0.173	0.088	0.322
Kerala	0.029	0.020	0.043	1.166	0.862	1.485
Madhya Pradesh	0.048	0.033	0.061	1.225	0.920	1.420
Maharashtra	0.024	0.014	0.040	0.664	0.423	0.894
Manipur	0.027	0.017	0.040	0.733	0.472	0.979
Meghalaya	0.057	0.049	0.067	1.821	1.521	2.450
Mizoram	0.062	0.047	0.075	2.972	2.245	3.841
Nagaland	0.053	0.044	0.060	2.216	1.942	2.579
Odisha	0.042	0.034	0.051	0.997	0.851	1.187
Puducherry	0.070	0.043	0.103	0.881	0.589	1.095
Punjab	0.033	0.024	0.045	1.234	0.934	1.563
Rajasthan	0.042	0.031	0.050	2.255	1.882	2.746
Sikkim	0.055	0.030	0.100	3.500	1.566	5.144
Tamil Nadu	0.016	0.008	0.024	0.595	0.315	0.866
Telangana	0.019	0.009	0.031	0.525	0.254	0.768
Tripura	0.153	0.136	0.169	12.914	9.612	16.666
Uttar Pradesh	0.040	0.030	0.049	1.238	1.042	1.503
Uttarakhand	0.041	0.032	0.052	1.167	0.952	1.554
West Bengal	0.050	0.034	0.064	2.110	1.738	2.545
India	0.029	0.022	0.037	0.881	0.754	1.014

Fig 2

Incidence-prevalence ratio by States in India, 2000–2019.

The years (2000–2019) are reflected on X-axis while IPR is on Y-axis. The highest bound on Y-axis is 0.250 for all except for Gujarat, Haryana, Madhya Pradesh, Meghalaya, Mizoram, Nagaland, Odisha and Punjab (0.500). Green line depicts target values of IPR for epidemic control, blue line represents point estimate and light blue shaded areas represents the uncertainty bounds for IPR of India/State for the period 2000–2019.

Fig 3

Incidence-mortality ratio by States in India, 2000–2019.

The years (2000–2019) are reflected on X-axis while IMR is on Y-axis. The highest bound on Y-axis is 5.000 for all except for Andhra Pradesh (7.000), Bihar (6.000), Delhi (6.000), Gujarat (12.000), Haryana (14.000), Meghalaya (12.000), Madhya Pradesh (14.000), Mizoram (45.000), Nagaland (14.000), Odisha (12.000) and Punjab (14.000). Green line depicts target value of IMR for epidemic control, blue line represents point estimate and light blue shaded areas represents the uncertainty bounds for IMR of India/State for the period 2000–2019.

Overall, 21 States/ UTs had IPR of more than 0.03 in 2019. This inlcuded seven States in the north-eastern region of India including Arunachal Pradesh [0.100, 0.061–0.122], Assam [0.059, 0.049–0.073], Meghalaya [0.057, 0.049–0.067], Mizoram [0.062, 0.047–0.075], Nagaland [0.053, 0.044–0.060], Sikkim [0.055, 0.030–0.100] and Tripura [0.153, 0.136–0.169]. IPR of less than 0.03 was noted in five States including Andhra Pradesh [0.009, 0.004–0.017], Karnataka [0.006, 0.003–0.011] and Tamil Nadu [0.016, 0.008–0.024] in southern parts of India. Rest of the States/UTs had IPR with uncertainty overlapping 0.03.

In Andhra Pradesh, IPR declined from 0.125 [0.069–0.186] in 2000 to 0.016 [0.010–0.023] in 2010 to 0.009 [0.004–0.017] in 2019. In Karnataka, IPR declined from 0.125 [0.063–0.183] in 2000 to 0.015 [0.009–0.022] in 2010 to 0.006 [0.003–0.011] in 2019. In Tamil Nadu, IPR declined from 0.054 [0.042–0.118] in 2000 through 0.024 [0.015–0.033] in 2010 to 0.016 [0.008–0.024] in 2019.

IMR was less than 1 in 2019 in seven States including that of Andhra Pradesh [0.199, 0.099–0.338], Karnataka [0.173, 0.088–0.322], Maharasthra [0.664, 0.423–0.894], Tamil Nadu [0.595, 0.315–0.866] and Telangana [0.525, 0.254–0.768]. In Andhra Pradesh, IMR was at 3.104 [1.267–6.115] in 2000, 0.192 [0.119–0.281] in 2010 and 0.199 [0.099–0.338] in 2019. In Karnataka, IMR was at 2.727 [1.043–4.675] in 2000, 0.176 [0.100–0.264] in 2010 and 0.173 [0.088–0.322] in 2019. In Maharasthra, IMR declined to 0.253 [0.165–0.390] in 2006 and then increased to 0.664 [0.423–0.894] in 2019. In Tamil Nadu, IMR was at 0.865 [0.477–3.508] in 2000, 0.281 [0.211–0.365] in 2005, 0.374 [0.228–0.636] in 2010 and 0.595 [0.315–0.866] in 2019. In Telangana, IMR declined to 0.325 [0.223–0.413] in 2007 and was at 0.525 [0.254–0.768] in 2019.

Eighteen States/UTs (Arunachal Pradesh, Assam, Bihar, Chhattisgarh, Delhi, Gujarat, Jharkhand, Jammu & Kashmir, Meghalaya, Mizoram, Nagaland, Rajasthan, Sikkim, Tripura, Uttar Pradesh, West Bengal, Dadra & Nagar Haveli and Daman & Diu) had estimated IMR of more than 1 in 2019. Rest of the States/UTs had IMR with uncertainty overlapping 1 in 2019.

The initial declining trend followed by stabilization or upward trend in IMR, as observed at the national level, was also noted in some of the States/ UTs including Chhattisgarh, Gujarat, Haryana, Manipur, Madhya Pradesh, Mizoram, Nagaland, Odisha, Punjab and West Bengal.

Discussions

The use of epidemiological metrices to assess trajectories of the HIV/AIDS epidemic and highlight areas for intervention is not new. The incidence:prevalence ratio has been used in the United Kingdom, United States of America, Denmark, Norway and Sweden in the past to estimate the HIV transmission dynamics [18-20]. These dynamic metrics are increasingly used since 2015 to assess the status of HIV epidemic control because they are rooted in sound epidemiological principles and analysis [4, 7–9, 21, 22]. These are very useful metrics to assess whether the direction of response to the epidemic is ‘on-track’ in the context of reducing the size of the epidemic and whether the epidemic is downsizing. However, the application of IPR and IMR to measure the Indian HIV epidemic transition has not been demonstrated earlier. This paper analyses the data from HIV estimations 2019 in India by its State/ UTs and presents the level and trends of IPR and IMR at the national and sub-national levels.

The significant decline in annual new HIV infections and AIDS-related deaths with continued low adult HIV prevalence has been documented in India [11, 13]. With a declining trend in IPR nationallly, the current analysis corraborates that HIV/AIDS epidemic control in India is on track. Yet, with an IPR of 0.029 [0.022–0.037] in 2019, programme intensity need to be maintained and augmented to attain and sustain the target IPR of <0.03 conclusively to achieve the epidemic control.

Nationally, the incidencce: mortality ratio in India is increasing since 2007. Given the trend, the IMR may surpass the threshold value of 1 in near future and the overall size of the people living with HIV/AIDS in India may also increase. This is consistent with a strong scale-up in the uptake of antiretroviral therapy (ART) leading to a very rapid decline in annual AIDS-related mortality [23].

In three southern states (Andhra Pradesh, karnataka and Tamil Nadu) both IPR and IMR is conclusively less than the target threshold in 2019. These three are among the states where historically the HIV epidemic was much higher than in the rest of India and hence the national AIDS response has focussed on these states since the early days of initiation of NACP with 70% or more of the estimated PLHIV are already on ART [23-26]. The response to the HIV epidemic in these states is on track and the total size of the PLHIV is expected to continue to decline in near future.

Twenty-one States/UTs across the country had IPR of more than 0.03 in 2019. Among these, sixteen (including seven States of Arunachal Pradesh, Assam, Meghalaya, Mizoram, Nagaland, Sikkim and Tripura in north-eastern India) also had IMR of more than 1 in 2019. These States need to be focussed under the programme as the PLHIV size in these States/UTs will grow over time with epidemic not in a state of control.

The current progress on the HIV epidemic transition threshold in India is consistent with the progress seen globally. The global incidence: prevalence ratio was 0.04 in 2019 with 25 countries having achieved the milestone of 0.03. Similarly, the threshold of <1 for IMR is expected to be achieved by a few countries with large HIV epidemics [27].

Both IPR and IMR are dynamic measures with roots in epidemiological theory about the sustainability of the transmission, but still have certain limitations [4–8, 23, 28, 29]. IPR threshold value of 0.03 for epidemic control assumes average survival of 33 years after HIV infection. The average survival may differ by regions and country based on the ART uptake, adherence etc which will have an impact on the IPR threshold value. The IPR threshold value to be asssumed for epidemic control in the context of ART uptake may be one of the factors while deliberating the threshold and behaviour of IPR [4, 7].

IPR and IMR work well when applied to the national or state population as a whole but are not suitable for the population subgroups where HIV acquisition and transmission is not be limited within the index subgroup. In India, where the HIV epidemic is concentrated with HIV prevalence among female sex workers, men who have sex with men, transgender people, and people who inject drugs are 7–28 times that of overall adult prevalence, this may lead to a false sense of complacency. Also, IMR may go down below the threshold level of 1 in settings with high AIDS-related mortality, as indicated in the case of states like Manipur, and thus be fallacious if seen in isolation in locations with low ART coverage.

Further, the scale-up of ART therapy reduces mortality, which shrinks the denominator and may create an upward trend in IMR even if incidence is decreasing and epidemic response is on track. The upward trend in IMR after 2007, seen at the country-level in India, is a resultant of this dynamics. Further, these measures do not reflect the status of legal, policy and social enablers and thus ignore the critical structural issues in AIDS response. Still, these two ratios provide critical insights into the current and future status of the epidemic in geographies as a whole by establishing if the epidemic is expanding or shrinking.

Our study limitations are also from source data. The study has used modelled estimates. There are inherent limitations to estimations based on modelling [4, 30]. The quality of modelled estimates depends on the quality of empirical data used as input data to inform the programme coverage and prevalence. The wider uncertainty bounds in 2019 estimates had limited capacity of this analysis to draw conclusive inference. The uncertainty bounds of estimates are usually influenced by aspects like quantity of surveillance data and use of population-based survey. Overall, the quanity and quality of sentinel surveillance in India has been described as good [31]. Also, prevalence estimates from population-based survey have been used in India for HIV burden estimation. The wider uncertainty bounds may be the outcome of the fact the State/UT-wise model is prepared during the HIV burden estimation process in India. Still, there is a need for examining the wider uncertainty bounds noted in modelled estimates.

The current study is based on analysis of outputs of a mathematical modeeling process recommended by UNAIDS. While modelled estimates on various epidemiological indicatiors like prevention of new HIV infections and AIDS-related mortality using globally used model are accepeted under NACP, there is a need for validation of modelled estimates of incidence and mortality. Investment in the components of second-generation surveillance focussing on incidence, mortality and case-based surveillance will triangulate the modelled estimates vis-à-vis empirical evidence and finally lead to a more reliable assessment on progress on the HIV epidemic transition threshold.

Despite the limitations, our study, the first to quantify the progress on epidemiological metrices by State and UTs in India to the best of our knowledge, highlighting potential challenges for national AIDS response. The overall progress at the national level masks the sub-national heterogeneity where the epidemic is expanding. To ensure that epidemic control is truly realised as envisioned in the 2030 SDGs, the interventions need to be tailored, expanded and intensified in the Indian States/ UTs, especially on prevention aspects, to reach the defined metric’s benchmark corresponding to epidemic transition.

Annual new HIV infections, annual all-cause mortality among PLHIV and total number of PLHIV (in 100,000) by States/UTs in India, 1990–2019.

(PDF)

Click here for additional data file.

IPR and IMR by States/UT in India, 1990–2019.

(PDF)

Click here for additional data file.

Transfer Alert

This paper was transferred from another journal. As a result, its full editorial history (including decision letters, peer reviews and author responses) may not be present.

9 Feb 2022

20 Apr 2022

Response to the observations of reviewers and suggestions of editors in the context of the manuscript ID “[PONE-D-21-34403] - [EMID:49e8df8c235ee894]” titled "Diversity in HIV Epidemic transitions in India: An application of HIV epidemiological metrices and benchmarks"

***

Response to the comments from Reviewer and Editor

We have gone through the observations of the reviewer and editor for our manuscript ID “[PONE-D-21-34403] - [EMID:49e8df8c235ee894]” titled "Diversity in HIV Epidemic transitions in India: An application of HIV epidemiological metrices and benchmarks".

We thank the reviewers for their constructive comments and we believe that we are now able to provide more clarity to the readers.

We have provided a detailed point-by-point response to all observations/comments (reviewer’s/editor’s comments in black, our replies in blue). Line numbering refers to the revised manuscript.

Reviewer’s Observations/Comments and Authors’ Response

1. General Observation/Comment

I’d like to thank the authors for the opportunity to review their manuscript. This is an important topic that is well explored by the paper. I recommend it for publication with a few suggested improvements/comments.

Reply: We sincerely thank the reviewer for appreciation of our work.

2. Technical Comments

(i) When discussing disease control, the ideal metric to be monitored is Rt. IPR and IMR are important proxies that help measure the growth or reduction of disease in a population. I think the authors do a good job of explaining why we should care about these metrics and how they relate to epidemic control. However, I think it would be useful to add information about their limitations.

Reply: We thank the reviewer for a clear guidance. We have added the limitations in the revised manuscript [line number ‘244’ to ‘266’].

(ii) The relation between IPR, epidemic control and the .03 target assumes that secondary infections are equally likely to occur at any time during the (on average) 30 years of time an individual is infected. While a useful assumption, this is in general not the case. Scale up of test and treat has reduced the infectivity of diagnosed individuals so that long term infections are less likely to transmit than short term ones. The result is that epidemic control may be reached at much higher than .03 IPR.

Reply: We agree with the reviewer. We have added this context in the revised manuscript [lines 244-251].

(iii) For IMR, changing the time between infection and death can create trends that have little to do with epidemic control. I believe you see this in the India data. The scale-up of ART therapy reduces mortality, which shrinks the denominator. This can create an upward trend in IMR even if incidence is decreasing. I interpret the upward trend after 2005 as an artifact of this.

Reply: We thank the reviewer for this insight. Although we had alluded to this point in paragraph 2 of the discussion section in our original manuscript, we have now included this additional perspective in the limitations of IMR in high ART coverage settings [lines 260-266 of the revised manuscript].

3. Other minor points

(a) If you can, it would be great if the images could be as high of quality as possible. I found myself zooming in on the individual graphs and they were readable, but could get a bit pixelated.

Reply: In pursuance of the suggestion of the reviewer, we have now updated the legends of the graphs clearly stating which colour of the line graph represents which indicator for the ease of understanding of the reader. The legend has been also updated to reflect the items on the X and Y-axis.

(b) 95-103: These equations are unreadable. It took me a while to realize that they were fractions. I thought they were formatting errors.

Reply: We thank the reviewer for this valuable suggestion. We have now presented the equations in an improved format by using the insert equation function. The revised equations are on lines 138 and 140.

(c) 164: This is too speculative. It may happen and the data may support continued reduction, but we shouldn’t just expect that HIV will go away.

Reply: We fully agree with the reviewer. We have edited the statement appropriately [line 217 of the revised manuscript].

(d) 119: I don’t see supplementals. Did you mean figures?

Reply: We had added the supplement at the time of original submission, but perhaps it was not accessible. We have uploaded the supplement again.

4. Grammer

Consider doing a close read through for grammar. Here are a few things I noticed.

a. 7: remove second “on”

b. 43: doesnot —> does not

c. 43: add “in” after “factor”

d. 59: delete “will”

e. 68: delete “.”

f. 69: Method —> Methods

g. 71: states —> state

Reply: We thank the reviewer and made changes as suggested. We have carefully reviewed our revised manuscript.

1. General Observation/Comment

I’d like to thank the authors for the opportunity to review their manuscript. This is an important topic that is well explored by the paper. I recommend it for publication with a few suggested improvements/comments.

Reply: We sincerely thank the reviewer for appreciation of our work.

2. Technical Comments

(i) When discussing disease control, the ideal metric to be monitored is Rt. IPR and IMR are important proxies that help measure the growth or reduction of disease in a population. I think the authors do a good job of explaining why we should care about these metrics and how they relate to epidemic control. However, I think it would be useful to add information about their limitations.

Reply: We thank the reviewer for a clear guidance. We have added the limitations in the revised manuscript [line number ‘244’ to ‘266’].

(ii) The relation between IPR, epidemic control and the .03 target assumes that secondary infections are equally likely to occur at any time during the (on average) 30 years of time an individual is infected. While a useful assumption, this is in general not the case. Scale up of test and treat has reduced the infectivity of diagnosed individuals so that long term infections are less likely to transmit than short term ones. The result is that epidemic control may be reached at much higher than .03 IPR.

Reply: We agree with the reviewer. We have added this context in the revised manuscript [lines 244-251].

(iii) For IMR, changing the time between infection and death can create trends that have little to do with epidemic control. I believe you see this in the India data. The scale-up of ART therapy reduces mortality, which shrinks the denominator. This can create an upward trend in IMR even if incidence is decreasing. I interpret the upward trend after 2005 as an artifact of this.

Reply: We thank the reviewer for this insight. Although we had alluded to this point in paragraph 2 of the discussion section in our original manuscript, we have now included this additional perspective in the limitations of IMR in high ART coverage settings [lines 260-266 of the revised manuscript].

3. Other minor points

(a) If you can, it would be great if the images could be as high of quality as possible. I found myself zooming in on the individual graphs and they were readable, but could get a bit pixelated.

Reply: In pursuance of the suggestion of the reviewer, we have now updated the legends of the graphs clearly stating which colour of the line graph represents which indicator for the ease of understanding of the reader. The legend has been also updated to reflect the items on the X and Y-axis.

(b) 95-103: These equations are unreadable. It took me a while to realize that they were fractions. I thought they were formatting errors.

Reply: We thank the reviewer for this valuable suggestion. We have now presented the equations in an improved format by using the insert equation function. The revised equations are on lines 138 and 140.

(c) 164: This is too speculative. It may happen and the data may support continued reduction, but we shouldn’t just expect that HIV will go away.

Reply: We fully agree with the reviewer. We have edited the statement appropriately [line 217 of the revised manuscript].

(d) 119: I don’t see supplementals. Did you mean figures?

Reply: We had added the supplement at the time of original submission, but perhaps it was not accessible. We have uploaded the supplement again.

4. Grammer

Consider doing a close read through for grammar. Here are a few things I noticed.

a. 7: remove second “on”

b. 43: doesnot —> does not

c. 43: add “in” after “factor”

d. 59: delete “will”

e. 68: delete “.”

f. 69: Method —> Methods

g. 71: states —> state

Reply: We thank the reviewer and made changes as suggested. We have carefully reviewed our revised manuscript.

Editor’s Observations/Comments and Authors’ Response

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming.

Reply: We thank the editor for the suggestion. We have updated the manuscript to meet PLOS ONE's style requirements to the best of our understanding.

2. You indicated that ethical approval was not necessary for your study. We understand that the framework for ethical oversight requirements for studies of this type may differ depending on the setting and we would appreciate some further clarification regarding your research. Could you please provide further details on why your study is exempt from the need for approval and confirmation from your institutional review board or research ethics committee (e.g., in the form of a letter or email correspondence) that ethics review was not necessary for this study? Please include a copy of the correspondence as an ""Other"" file.

Reply: We thank the editor for the observation. We wish to clarify that HIV Surveillance and Estimation is an integral part of the spectrum of activities of the National AIDS Control Organization (NACO) of the Government of India. This activity is periodically conducted using a specially created programmatic framework built on the foundation of government institutes and partners. As part of the surveillance activity, when primary data and sample collection is done from the survey participants, informed consent is taken from them in alignment with national guidelines. The institutions involved in primary data collection routinely submit their proposals for the surveillance program to their respective ethics committees to seek approval and the survey at each site is initiated only after the local ethics committee approves the proposal. However, utilizing this program generated data for policy making and programmatic improvement is a mandate of the National AIDS Control Program. We have attached the Ethics Committee approval for the activity of integrated bio-behavioural surveillance and HIV sentinel surveillance for reference as “Other” file.

HIV burden estimation is carried out after each round of Surveillance by NACO by making use of aggregated de-identified data. This work has been published periodically in the past. It may be noted that primary data collection from human subjects is not part of HIV burden estimation exercise. The current manuscript is based on the analysis of aggregated de-identified outputs generated through the HIV Estimations 2019 model for each of the Indian State or Union Territory.

In summary, ethics committee approvals have been taken by all the agencies that contribute primary data to HIV surveillance program, but state level HIV burden estimations have been done as part of mandate of the National AIDS Control Program.

3. Please amend your authorship list in your manuscript file to include all author

Reply: We thank the editor for the suggestion. We have updated the manuscript file to include all author [line 4-19].

4. In your Data Availability statement, you have not specified where the minimal data set underlying the results described in your manuscript can be found. PLOS defines a study's minimal data set as the underlying data used to reach the conclusions drawn in the manuscript and any additional data required to replicate the reported study findings in their entirety. All PLOS journals require that the minimal data set be made fully available.

Reply: The minimal data set underlying the results described in the manuscript can be found in the supplementary tables. Additional de-identified data can be made available upon request to the researchers who meet the criteria for NACO’s data-sharing guidelines. Interested researchers should write to Dr Chinmoyee Das, [c.das@gov.in], Head of Division, Strategic Information (NACO, MoHFW, Govt of India).

5. Please review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript. If you need to cite a retracted article, indicate the article’s retracted status in the References list and also include a citation and full reference for the retraction notice.

Reply: We have reviewed the references and ensured that they have been correctly cited. To the best of our knowledge, we have not cited any papers that have been retracted.

Submitted filename: Authors response to reviewers comments .docx

Click here for additional data file.

23 May 2022

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Nicky McCreesh

Academic Editor

PLOS ONE

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #2: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #2: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #2: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #2: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #2: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #2: (No Response)

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #2: Yes: John Stover

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

5 Jun 2022

Academic Editor Observations/Comments and Authors’ Response

1. Comment 1:

The data source used, 'India HIV Estimates 2019' shows very large amounts of uncertainty in the estimates of numbers of PLHIV, HIV incidence, and HIV mortality. This uncertainty needs to be carried through into the estimates presented in the manuscript, as the results cannot be correctly interpreted without confidence intervals. The level of uncertainty in the results should also be reflected in the interpretation of the results and the language used throughout the manuscript. For instance, statements such as "Ten States/ UTs of Andaman ... had estimated IMR of one or less." should not be made if the confidence intervals overlap one.

Reply:

We agree that we have provided only the point estimates for IPR and IMR. We concur with the Editor observations of providing uncertainty bounds. Adding uncertainty bounds on IPR and IMR will not only increase the scientific rigour of the manuscript but also improve inferences as noted by the Editor.

The current manuscript is based on the outputs generated using Spectrum Model under 2019 round of HIV burden estimates in India. The Spectrum model calculates uncertainty bounds around each estimate including that on new infections and deaths among PLHIV. However, we were of the opinion that using lower bounds or upper bounds of new infections, deaths and PLHIV to create uncertainty bound around IPR and IMR may not be the right approach as these would be applied for both numerator and denominator.

We noted that UNAIDS publications have provided IPR with uncertainty bounds though method for the same was not described. And hence we reached to Dr Keith Sabin (Strategic Information Department, UNAIDS, Geneva, Switzerland) to understand the calculation of uncertainty bounds. Dr Sabin informed that proportional uncertainty around the new infections is used to calculate the uncertainty bound around IPR and IMR. To be consistent with globally used methodology, we applied the same for our manuscript. This method has been stated in Line Number 136-138 of the revised manuscript.

The Results section has been updated to include the uncertainty bounds around tables and figures. The table 1, beginning at Line Number 177, now present State/UT-wise IPR and IMR with uncertainty bounds. Besides all the relevant paragraphs (Line Number 159-165, Line Number 180-200) in the results sections have results with uncertainty bounds.

We have revised the inferences and language of the discussions in view of the insights provided by uncertainty bounds. The revision is reflected in Line Number 208 to 239 of the revised manuscript.

We have noted the wider uncertainty bounds as one of the limitations (Line Number 270-277).

2. Comment 2:

The new paragraph starting from line 244 is not correct. I appreciate that it was added in response to a suggestion made by a reviewer, but it is not necessary to make changes suggested by reviewers if they are not factually correct. The prevalence of HIV will grow in a population if the rate of new infections is higher than the rate at which people are removed from the pool of prevalence infections. The time of transmission relative to the duration of disease does not effect the interpretation of the IPR in such a simplistic way. I think the mistake the reviewer was making was in thinking that if people mostly only transmit in the first 5 years following infection, the the IPR at which control is achieved should be 1/5, not 1/30. What that misses is that in that case only prevalent infections in the first 5 years following infection should be included in the numerator for the IPR, and therefore the two factors will cancel out.

Reply:

We do see the point mentioned by the academic editor. In fact, we reviewed the UNAIDS report titled ‘MAKING THE END OF AIDS REAL: CONSENSUS BUILDING AROUND WHAT WE MEAN BY EPIDEMIC CONTROL’ and paper titled "Epidemiological metrics and benchmarks for a transition in the HIV epidemic’ by Ghys, Peter D., et al. We noted that both of these refers to threshold of 0.03 as more of a ‘rule of thumb’ and indicated for more discussions around the threshold. As ART uptake has significant impact on transmission of new infections, we believe that this is area that may be included as one of the points in future discussions on threshold value and behaviour of IPR.

In view of above, we have retained the reference to potential impact of ART on IPR threshold. However, we have removed the sentence ‘As a result, the epidemic control may be reached at much higher than 0.03 IPR’ to avoid any comments on the threshold level of epidemic control in our manuscript. We have added a sentence ‘Impact of ART uptake on the IPR may be one of the factors while deliberating the threshold and behaviour of IPR’ (Line Number

Submitted filename: Response to Reviewers.docx

Click here for additional data file.

7 Jun 2022

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Nicky McCreesh

Academic Editor

PLOS ONE

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

18 Jun 2022

PONE-D-21-34403R2

Diversity in HIV Epidemic transitions in India: An application of HIV epidemiological metrices and benchmarks

PLOS ONE

***

Response to Reviewers/Editor

We have gone through the observations of the reviewer and editor for our manuscript ID “PONE-D-21-34403R2” titled "Diversity in HIV Epidemic transitions in India: An application of HIV epidemiological metrices and benchmarks".

We thank the Academic Editor for very constructive comments. We believe the review has enhanced the rigour of the manuscript.

We have provided a detailed point-by-point response to all observations/comments (editor’s comments in black, our replies in blue). Line numbering refers to the revised manuscript.

***

Academic Editor Observations/Comments and Authors’ Response

1. Comment 1:

Thanks for adding in uncertainty estimates, they greatly increase the potential usefulness and applicability of the work. Unfortunately, the way they are calculated is not quite correct (based on the way it is described in the methods). It takes into account the uncertainty resulting from the uncertainty in the incidence estimates, but not the uncertainty resulting from the mortality and prevalence estimates. This could mean that the uncertainty is underestimated. Alternatively, if the incidence and mortality/prevalence estimates are correlated between model runs, it may mean that the uncertainty is overestimated.

Reply:

We agree with the comment.

2. Comment 2:

I have not used the Spectrum models myself. Is it possible to export the estimates from individual model runs, using the uncertainty analysis tool? If it is possible, the simplest way to obtain better estimates of the uncertainty would be to divide the incidence by the mortality/prevalence estimate separately for each individual model run, and then to obtain the 2.5% and 97.5% percentiles for the ratios to give 95% uncertainty intervals. Please make sure that the uncertainty intervals that this approach generates seem reasonable to you however, it is possible that the way that the Spectrum model is parameterised and calibrated would mean that this approach would generate unreasonably narrow uncertainty intervals.

Is this is not possible, then there may not be any better approach than the one used, but please discuss the limitations of the way that the uncertainty intervals are generated in the Discussion section.

Reply:

We thank the Academic Editor for an extremely useful suggestion. The ‘Uncertainty Analysis’ tool in Spectrum indeed generates plausible range by running 1000 iterations. Results of each of the iteration for key indicators may be exported. Given this provision, we revised the method for calculating the uncertainty bounds. The revision has been suitably reflected under Method [Line Number 120-124, 145-153 of revised manuscript], Results [Line Number 170-175, 194-223 of revised manuscript] and Discussions [Line Number 240, 243-245, 249-261 of revised manuscript].

3. Comment 3:

I agree entirely that the threshold of 0.03 is a rule of thumb only, and feel that a paragraph discussing this is important. What I do not understand is this sentence: “the 0.03 target assumes that secondary infections are equally likely to occur at any time during the (on average) 30 years of time an individual is infected”. As far as I can see, the references you provided in your response say nothing about the target assuming this, and I do not understand how it does. Please explain this statement and/or provide a reference for it, or remove it. It could be replaced with a brief discussion of the major assumption made in setting the threshold: that people live with HIV for a mean of 30 years.

Reply:

We have revised the paragraph suitably [Line Number 268-272 of revised manuscript].

Submitted filename: Response to Reviewers.docx

Click here for additional data file.

21 Jun 2022

Diversity in HIV Epidemic transitions in India: An application of HIV epidemiological metrices and benchmarks

PONE-D-21-34403R3

Dear Dr. Kumar,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Nicky McCreesh

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

4 Jul 2022

PONE-D-21-34403R3

Diversity in HIV Epidemic transitions in India: An application of HIV epidemiological metrices and benchmarks

Dear Dr. Kumar:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Nicky McCreesh

Academic Editor

PLOS ONE