Journal of Pharmaceutical Care & Health Systems

Journal of Pharmaceutical Care & Health Systems
Open Access

ISSN: 2376-0419

+44 1300 500008

Research Article - (2018) Volume 5, Issue 3

An Interrupted Time Series Analysis of the Second line Antiretroviral Policy Change from Lopinavir Boosted with Ritonavir to Atazanavir Boosted with Ritonavir Based Regimens in Namibia

Babafunso A. Adenuga1*, Dan Kibuule1,2 and Kayode DS Bamitale2
1Department of Pharmacy Practice & Policy, Faculty of Health Sciences, School of Pharmacy, University of Namibia, Windhoek, Namibia
2Department of Pharmacology and Therapeutics, Faculty of Health Sciences, School of Pharmacy, University of Namibia, Windhoek, Namibia
*Corresponding Author: Babafunso A. Adenuga, Department of Pharmacy Practice & Policy, Faculty of Health Sciences, School of Pharmacy, University of Namibia, Windhoek, Namibia, Tel: +264(0)818659993 Email:

Keywords: Policy change; Second line regimens; Antiretroviral therapy; Guidelines; Patients

Introduction

Over the last two decades, the treatment of people living with HIV/ AIDS (PLWHA) has been evolving with different classes of drugs being introduced, often fast-tracked for approval [1,2]. The goal of therapy has been to improve the health outcomes of patients and allow them to integrate and be useful in the society in which they found themselves.

World Health Organization (WHO) introduced guidelines aimed at assisting low and middle income countries (LMIC) to adopt viable therapy regimens for PLWHA. In the 2016 WHO Guidelines on the Use of Antiretroviral drugs [3], atazanavir boosted with ritonavir was introduced as the preferred Protease Inhibitor (PI) backbone of second line regimens, while lopinavir boosted with ritonavir was the alternate PI. LMICs can adopt this recommendation when formulating or updating their guidelines. As at 2015, Clinton Health Access Initiative (CHAI) market estimated about 22% of adults on second line regimens were taking ATV/r as their PI backbone [4].

According to the WHO, 2.9% of patients on ARV were on second line regimens with LPV/r as the PI backbone, ATV/r is being used as an alternate to LPV/r [3]. Namibia has about 210 000 PLWHA; out of these, about 150 000 patients are on antiretroviral therapy (ART). Among the patients on ART in Namibia, about 5% are currently on second line regimens [5]. This percentage is higher than the 2.9% reported by the WHO (2010), however, it is believed that the number will go up over time.

First line regimens such as the Nucleoside Reverse Transcriptase Inhibitors (NRTIs) e.g. Zidovudine and Lamivudine and Non- Nucleoside Reverse Transcriptase Inhibitors (NNRTIs) e.g. Efavirenz and Nevirapine, used as part of antiretroviral treatment (ART) regimens are known to have low genetic barriers, thus, they develop resistance to the drugs earlier compared to the second line regimens [6]. The implication of this is, patients that are being maintained on first line regimens will experience treatment failure due to development of resistance to the ARV over time [7].

The Ministry of Health and Social Services, Namibia introduced a new ART guideline in August 2014 [8]. It included the introduction of ATV/r as a preferred Protease inhibitor (PI) backbone for second line regimens in Namibia to replace LPV/r. Atazanavir is known to be safer compared to Lopinavir with regard to metabolic adverse effects such fat redistribution [9]. Thus, the introduction of the drug assumes a better compliance and treatment outcomes for patients that were previously on LPV/r based regimens.

Objective

The study sought to assess the implementation of the new ART guidelines with specific emphasis on the time to introduction of ATV/r as a PI backbone for second line regimens in Namibia. Also, to estimate the projected number of patients will remain on LPV/r based regimens by January 2019 as mentioned in Table 1 and Figure 1.

pharmaceutical-care-health-systems-Time-series

Figure 1: Time series plot of switching LPV/r to ATV/r based regimens in 6 month intervals.

t Year Date LPVr MA/LPVr (2) CMA/LPVr (2) St, It St Deseasonalise Yt/St Tt Prediction
1 Y1 Jan-15 3937 - - - 1 3930 4550 4558
2 Feb-15 4217 4077 4084 1.03 0.99 4245 4537 4507
3 Mar-15 3964 4091 4084 0.97 1 3953 4485 4497
4 Apr-15 4189 4077 4124 1.02 1 4173 4511 4528
5 May-15 4152 4171 4195 0.99 0.99 4184 4498 4463
6 Jun-15 4286 4219 4211 1.02 1 4296 4485 4474
7 Jul-15 4119 4203 4260 0.97 1.01 4073 4472 4522
8 Aug-15 4517 4318 4393 1.03 0.99 4565 4458 4412
9 Sep-15 4417 4467 4443 0.99 1 4425 4445 4437
10 Oct-15 4419 4418 4488 0.98 1 4401 4432 4450
11 Nov-15 4695 4557 4641 1.01 1 4694 4419 4419
12 Dec-15 4756 4726 4772 1 1 4778 4406 4385
13 Y2 Jan-16 4880 4818 4863 1 0.99 4927 4393 4351
14 Feb-16 4935 4908 5000 0.99 0.99 5006 4380 4317
15 Mar-16 5248 5092 5003 1.05 0.98 5350 4366 4283
16 Apr-16 4580 4914 4849 0.94 0.98 4692 4353 4249
17 May-16 4988 4784 4883 1.02 0.97 5135 4340 4216
18 Jun-16 4975 4982 4898 1.02 0.97 5147 4327 4182
19 Jul-16 4652 4814 4614 1.01 0.96 4837 4314 4149
20 Aug-16 4178 4415 4272 0.98 0.96 4366 4301 4115
21 Sep-16 4078 4128 4080 1 0.95 4283 4288 4082
22 Oct-16 3986 4032 3921 1.02 0.95 4208 4274 4049
23   Nov-16 3632 3809 3560 1.02 0.94 3854 4261 4016
24 Dec-16 2989 3311 3337 0.9 0.94 3188 4248 3984
25 Y3 Jan-17 3739 3364 3704 1.01 0.93 4008 4235 3951
26 Feb-17 4350 4045 4158 1.05 0.93 4687 4222 3918
27 Mar-17 4192 4271 4002 1.05 0.92 4540 4209 3886
28 Apr-17 3275 3734 3566 0.92 0.92 3566 4196 3854
29 May-17 3521 3398 3516 1 0.91 3853 4182 3822
30 Jun-17 3748 3635 3668 1.02 0.91 4124 4169 3790
31 Jul-17 3653 3701 3665 1 0.9 4040 4156 3758
32 Aug-17 3605 3629 3608 1 0.9 4009 4143 3726
33 Sep-17 3568 3587 3584 1 0.89 3989 4130 3694
34 Oct-17 3595 3582 3586 1 0.89 4041 4117 3663
35 Nov-17 3586 3591 3612 0.99 0.88 4052 4104 3631
36 Dec-17 3679 3633 3632 1.01 0.88 4180 4090 3600
37 Y4 Jan-18 3585 3632 3568 1 0.88 4096 4077 3569
38 Feb-18 3422 3504 - - 0.87 3931 4064 3538
39 Mar-18 3377 - - - 0.87 3901 4051 3507
40 Apr-18 - - - - 0.86 - 4038 3476
41 May-18 - - - - 0.86 - 4025 3446
42 Jun-18 - - - - 0.85 - 4012 3415
43 Jul-18 - - - - 0.85 - 3998 3385
44 Sep-18 - - - - 0.84 - 3985 3355
45 Oct-18 - - - - 0.84 - 3972 3324
46 Nov-18 - - - - 0.83 - 3959 3294
47 Dec-18 - - - - 0.83 - 3946 3265
48   Jan-19 - - - - 0.82 - 3933 3235

N.B: The data used and forecast results obtained are in red.

Table 1: Table showing predicted number of patients on LPV/r based second line regimens form April 2018 to January 2019.

Methods

Study design and setting

A cross-sectional interrupted time series analysis of all the national second line ART patients by regimen data from January 2015 to March 2018 in Namibia was conducted. Interrupted time series is a good study design when the impact of a new intervention in the public healthcare sector is being monitored [10-12]. Research ethics board approvals were obtained from MoHSS (Reference No. 17/3/3) and Ethics Committee of the University of Namibia (Reference No. SOPHA/209/2017). Patient informed consent was not required because only reported data were used in the study. This report adheres to the Reporting of studies Conducted by using the Observational Routinely-collected health Data statement.

Data sources

National second line ART Patients by Regimen data were retrieved from the MoHSS Pharmaceutical Management Information System Dashboard (PMIS Dashboard). ARV consumption data between the months of January 2015 to March 2018 were included in the study. The focus of the study was the adults on second line ARV regimens with LPV/r and ATV/r as the PI backbone. Adult patients’ data were used in this study, considering the ethical clearance granted for the study was only for adults. The data obtained were aggregated into months and entered on a spread sheet prior to exporting into R studio® and Minitab® (Free Edition) that were used in the analysis of the data.

Study population

Aggregated monthly national ART Patients by regimen data of adult patients were included in the study. Data for LPV/r based regimens between January 2015 and March 2018 and ATV/r based regimens data between January 2017 and March 2018. Adult patients of Namibian origin currently on second line regimens in the public health sector of Namibia were included. Data were retrieved from the MoHSS Pharmaceutical Information Dashboard. Demographics of the patients were not included or required for the study.

Outcome measures

Our primary outcome was the time to implementation of the new ART guideline, with particular reference to ATV/r introduction as a preferred second line regimen. Secondary outcome included inference of the number of patients that should be on second line regimens based on ATV/r, if all the patients were switched to the drug as envisaged by the new guideline.

Population characteristics

Adult patients, from 18 years upwards, of Namibian origin currently on second line regimens in the public health sector of Namibia were included in the study Figure 2. Demographics of the patients were not included or required for the study.

pharmaceutical-care-health-systems-Scatter-plot

Figure 2: Scatter plot of LPV/r and ATV/r based second line regimens consumption over time.

Statistical analysis

The data were summarised using Microsoft Excel spreadsheet, aggregating the number of patients on LPV/r and ATV/r regimens on a monthly basis. The data were exported to R and Minitab, the software used in data analysis. ATV/r based regimens were reported first in January 2017 in the PMIS Dashboard, however, LPV/r based regimens has been used as PI backbone for second line regimens as early as 2010 [13]. Regimen characteristics were compared at 2 time points [1] periintervention period January 2016; and [2] post-intervention August 2017. The two predefined time points were selected a priori to identify any potential variability or nonstable health system utilization patterns of the population. Interrupted time series analysis was performed to examine the impact of the ART Guideline August 2014 on each of our primary and secondary outcomes of interest. The models quantify the time, level and trend change following the intervention, while accounting for the autocorrelation of observations. The 2014 guideline introduction was selected as the intervention, as it was the first year that ATV/r was officially introduced and promoted in the Namibian public health system. The onset date for the intervention was lagged nearly 12 months (until January 1, 2016) to account for the delay that would be required for rollout of the new guideline, the training of healthcare workers responsible for the treatment of patients on the new guideline, and the new regimen in particular and for the procurement and distribution of the regimen by the Central Medical Stores (CMS). Each year was divided into monthly intervals, spanning a total of 39 months during the study period (2015-2018) as shown in Figure 3. The outcome variables were the number of patients on LPV/r and ATV/r based regimens and the time to implementation of the 2014 ART guideline. The following segmented regression model was used:

pharmaceutical-care-health-systems-series-analysis

Figure 3: Time series analysis of switching LPV/r to ATV/r based regimens.

Yt = β0 + β1*T+ β2*Xt + β3*T*Xt +et

Where, Yt is the outcome, i.e. the number of patients that are switched from LPV/r based regimens to ATV/r based regimens at time t, T is the time (in months) that elapsed since the start of the study, Xt is a dummy variable indicating the pre-intervention period (coded 0), or the post-intervention period (coded 1); β0 estimates the baseline outcome at T=0; β1 is an estimate of the peri-intervention outcome trend (i.e. the change in outcome with time); β2 is an estimate of the change in outcome immediately after the peri-intervention period, i.e. compared to the outcome at the pre-intervention period; β3 the change in the post-intervention outcome trend compared to the preintervention outcome trend; et represents the random variability not explained by the model. For all statistical tests, a p-value of ≤ 0.05 was considered to be significant.

MS Excel 2007 was used to forecast the number of patients that will be on either LPV/r or ATV/r based regimens per month.

Results

In Table 2, an average of 4336 patients were on LPV/r based second line regimens between January 2015 and December 2016, this varies considerably from month to month (range: 2989–5248), prior to the introduction and consumption report of ATV/r. This number decreased to an average of 3533 (range: 2708–4329) between January 2017 and March 2018. The number of patients switched to ATV/r based second line regimens increased over time with an average of 1704 (range: 14 – 3471) patients between January 2017 and March 2018. Table 2 displays a reduction in the number of patients on LPV/r based regimens over time, however, a projected total number of patients on second line regimens is depicted in Figure 4 at approximately 7000, if every patient were switched to ATV/r based second line regimens at the end of March 2018.

pharmaceutical-care-health-systems-Trend-analysis

Figure 4: Trend analysis plot for LPV/r showing a downward trend in the number of patients over time.

Date LPVr ATVr Time Level Trend
Jan-15 3937 0 1 0 0
Feb-15 4217 0 2 0 0
Mar-15 3964 0 3 0 0
Apr-15 4189 0 4 0 0
May-15 4152 0 5 0 0
Jun-15 4286 0 6 0 0
Jul-15 4119 0 7 0 0
Aug-15 4517 0 8 0 0
Sep-15 4417 0 9 0 0
Oct-15 4419 0 10 0 0
Nov-15 4695 0 11 0 0
Dec-15 4756 0 12 0 0
Jan-16 4880 0 13 1 1
Feb-16 4935 0 14 1 2
Mar-16 5248 0 15 1 3
Apr-16 4580 0 16 1 4
May-16 4988 0 17 1 5
Jun-16 4975 0 18 1 6
Jul-16 4652 0 19 1 7
Aug-16 4178 0 20 1 8
Sep-16 4078 0 21 1 9
Oct-16 3986 0 22 1 10
Nov-16 3632 0 23 1 11
Dec-16 2989 0 24 1 12
Jan-17 3739 31 25 1 13
Feb-17 4350 19 26 1 14
Mar-17 4192 1778 27 1 15
Apr-17 3275 2046 28 1 16
May-17 3521 2150 29 1 17
Jun-17 3748 1601 30 1 18
Jul-17 3653 2561 31 1 19
Aug-17 3605 2714 32 1 20
Sep-17 3568 2866 33 1 21
Oct-17 3595 3149 34 1 22
Nov-17 3586 3133 35 1 23
Dec-17 3679 3421 36 1 24
Jan-18 3585 3561 37 1 25
Feb-18 3422 3492 38 1 26
Mar-18 3377 3598 39 1 27

Table 2: ARV consumption by regimen showing the number of patients reported to be on each regimen by month.

Table 3 shows an increased level in the number of patients that were being switched to LPV/r based second line regimens before the implementation of the guideline with approximately 210 patients being initiated LPV/r based second line regimen every month (210.29; CI: -234.86, 655.45) however, the downward trend of the number of patients on the same regimen continued after the introduction of ATV/r based second line regimens, as can be seen on the graph. There was a downward trend in the number of patients enrolled on LPV/r based second line regimens (-127.22; CI -184.98, -69.46) monthly; this was observed by the spike in the number of patients switched to ATV/r based regimens monthly as shown in Figure 5.

pharmaceutical-care-health-systems-upward-trend

Figure 5: Trend analysis plot ATV/r showing an upward trend in the number of patients of patients over time.

Variable Parameter Estimate (95% CI) Standard Error t Value p-value
β0 (pre-intervention level) 3883.76 (3476.05, 4291.46) 200.83 19.339 <0.0001
Β1 (pre-intervention trend) 64.91 (9.51,120.30) 27.29 2.379 0.023
Β2 (pre-intervention level) 210.29 (-234.86, 655.45) 219.27 0.939 0.344
Β3 (pre-intervention trend change) -127.22 (-184.98, --69.46) 28.45 -4.471 <0.0001

p=0.05, f statistic=27.25

Table 3: Estimated coefficients for the interrupted time series analysis of the introduction of ATV/r based second line regimens.

The number of patients on LPV/r based second line regimens was predicted from April 2018 to January 2019. There was a decline over time, with the highest number of patients in April 2018 (3476) and lowest in January 2019 (3235).

Discussion

There was a slight increase in the number of patients before the implementation being initiated on LPV/r based second line regimens. There was an increase in the number of patients on LPV/r by about 210 patients immediately after the implementation, however, the trend went down by about 127 patients per month indicating that patients are being switched to ATV/r based second line regimens.

The Namibian ART guideline [8] does not describe the eligibility criteria to be used when switching patients to ATV/r based second line regimens. Prescribers assessed patients’ eligibility to be switched to the new regimen, ATV/r based on some criteria such as viral load suppression and being “stable” on their current regimens. The uptake of the new regimen was not rapid, as the first reported consumption was in January 2017 which was 16 months after the introduction of the new guideline. This has implications on the treatment outcomes of patients maintained on LPV/r based regimens or other second line regimens, due to the fact that ATV/r is known to be more tolerable and has less adverse reactions compared to LPV/r and NVP based second line regimens that were used prior to the introduction of ATV/r [13]. There were no competing interventions at or around the time of the introduction of the new guideline.

Considering the clinical work up necessary to switch a patient from one regimen to the other, the projected start date for ATV/r was not met, until January 2017 when the first reported usage of the regimen was documented. This has an impact on the adherence of patients who has been informed about a regimen that is much better than their current regimens. Patients tend to lose confidence in the system and invariably their healthcare providers who promised a “better life” for them.

It took about 28 months (August 2014 to December 2016) for the implementation of the ART guideline [8] with regard to the introduction of ATV/r as preferred PI back bone for second line regimens in Namibia. The delay in the introduction of ATV/r may be due to the communications within the MoHSS and with its partners; training of personnel such as medical doctors, pharmacists and nurses who are responsible for the treatment or switching of eligible patients; availability of the new drug at the Central Medical Stores prior to or immediately after the introduction of the new regimen. From the predicted number of patients over the next 12 months, it is evident that switching of all patients from LPV/r based second line regimens to the preferred ATV/r based second line regimens will take a long time, thus, the impact of such delay in policy implementation on patients’ health outcomes has to be determined after at least one year of policy implementation. Quantitative or qualitative research into prescribers’ ability to implement the policy or hindrances to implementation of the policy may be an area to be explored in the future.

Limitations

Underreporting by facilities is common, as seen in the fluctuating number of patients every month. Though, this may be expected due to the current manual reporting system that is used by the MoHSS.

Non-availability of updated reporting parameters for ATV/r on EDT/EPMS, prior to, during or after the introduction of ATV/r, thus, possibly hampers the reporting of the regimen consumption.

Reporting rates by the facilities that are required to submit monthly reports to the national level contributes to accuracy or completeness of reports available on the PMIS Dashboard.

Policy makers and implementers should work together in ensuring the success of policies that has far reaching impact on the healthcare system and welfare of patients and the nation as a whole.

A population level design was used in the study based on a national database. The impact of time varying covariates such as programmatic and economic impacts could not be accessed. Also, we could not assess patient-level and provider-level variables.

Also, having used secondary data from the national database, we could not validate its accuracy despite the quality control measures being put in place by the MoHSS.

Nonetheless, this study provides a true reflection of the current situation in Namibia and possibly, other LMICs.

Conclusion

The implementation of the August 2014 ART guideline incorporating the use of ATV/r as PI backbone did not take off immediately, but it took about 28 months before the first consumption reports were made. With the increasing number of patients that are failing first line regimens due to varying reasons, implementation of such guidelines should be given priority so as to minimize the morbidity or mortality that may occur due to late initiation/switching of patients that are failing the regimens they are on [14]. One observation that was made in the 2014 guideline was the absence of the eligibility criteria for patients to be switched to ATV/r. also, ATV/r based second line regimens were not categorically asserted as the preferred option of patients on second line ART, and however, this was addressed in the 5th edition of the ART guideline released in 2016 [15].

Recommendations

A formal system for implementation of guidelines in specified time frame, guided by the activities necessary to accomplish the goals of the guidelines needs to be set up.

Integrating the healthcare delivery reporting systems electronically will minimize underreporting and provide full picture of the ARV regimens used by patients and number patients on various regimen will be known. If reports are being generated simultaneously as patients are seen, it will minimize underreporting or non-reporting of activities performed or consumption data of ARVs.Having a program for the facilities for every service provided that feeds into the national EPMS program will enhance the actual quantification of the number of patients on different regimens at a particular time.

Ensuring that every stakeholder are carried along during implementation of guidelines is important, considering the rising number of patients that are failing first line regimens and those that are yet to be transitioned to the preferred ATV/r regimen [16].

Declarations

• Ethical approval was obtained from the Ethics Committee of the Ministry of Health and Social Services and Ethics Committee of the University of Namibia.Consent for publication - Not applicable

• The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

• There was no competing interest in the conduct of this study.

Funding

We had no external source of funding for this study.

• BAA did the write up and final data analyses. KD did the initial data analysis and review of the manuscript. TWR did the review and gave new insights to the write up. BKDS reviewed part of the manuscript.

• Acknowledgements - Wuletau Zeleke of the Department of Pharmaceutical Services, MoHSS, Namibia assisted with the data that was used for this study.

References

  1. Wilson WH, Schenkein DP, Jernigan CL, Woodcock J, Schilsky RL (2013) Reevaluating the accelerated approval process for oncology drugs. Clinical Cancer Res 19: 2804-2809.
  2. United States Food and Drug Administration (2018) Fast Track, Breakthrough Therapy, Accelerated Approval, Priority Review.
  3. WHO (2016) Consolidated guidelines on the use of antiretroviral drugs for treating and preventing HIV infection: recommendations for a public health approach. Geneva. World Health Organization.
  4. Clinton Health Access Initiative (2016) ARV Market Report: The State of the Antiretroviral Drug Market in Low- and Middle-Income Countries.
  5. Mackie N (2018) Resistance to non-nucleoside reverse transcriptase inhibitors. In A M Geretti, Antiretroviral Resistance in Clinical Practice. London.
  6. Gega A, Kozal JM (2011) New Technology to Detect Low-level Drug-resistant HIV Variants. Future Virology 6: 17-26.
  7. De Luca A (2006) The impact of resistance on viral fitness and its clinical implications. Antiretroviral Resistance in Clinical Practice PP: 1-15.
  8. Ministry of Health and Social Services (2014) National Guideline on Antiretroviral Therapy, 4th Edition, Windhoek, Namibia.
  9. Molina JM, Andrade VJ, Echevarria J, Chetchotisakd P, Corral J, et al. (2010) Once-daily atazanavir/ritonavir compared with twice-daily lopinavir/ritonavir, each in combination with tenofovir and emtricitabine, for management of antiretroviral-naïve HIV-1-infected patients patients: 96-week efficacy and safety results of the Castle Study. J Acquired Inmmune Deficiency Syndrome 53: 323-332.
  10. Bernal JL, Cummins S, Gasparrini A (2017) An Interrupted time series regression for the evaluation of public health interventions: a tutorial. Inter J Epidemiology PP: 348-355.
  11. Penfold RB, Zhang F (2013) Use of Interrupted Time Series Analysis in Evaluating Health Care Quality Improvements. Academic Paediatrics 13: S38-S44.
  12. Wagner AK, Soumerai SB, Zhang F, Ross DD (2002) Segmented regression analysis of interrupted time series studies in medication use research. J Clinical Pharmacy and Therapy 27: 299-309.
  13. Ministry of Health and Social Services (2010) National Guideline on Antiretroviral Therapy. 3rd Edition, Windhoek, Namibia
  14. Ministry of Health and Social Services (2016) National Guideline on Antiretroviral Therapy. 5th Edition, Windhoek, Namibia.
  15. Avert - Global information and education on HIV and AIDS (2017) Antiretroviral Treatment for HIV.
Citation: Adenuga BA, Kibuule D, Bamitale KDS (2018) An Interrupted Time Series Analysis of the Second line Antiretroviral Policy Change from Lopinavir Boosted with Ritonavir to Atazanavir Boosted with Ritonavir Based Regimens in Namibia. J Pharma Care Health Sys 5: 195.

Copyright: © 2018 Adenuga BA, et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Top