Inability-to-walk-unaided—a single WHO danger sign predicts in-hospital mortality in people with HIV under routine care conditions in a low-resource setting

Abstract

Background:

People with advanced HIV admitted to hospitals are at high risk of mortality. Serious illness can be identified using WHO-defined danger signs (“WHO score”) or bedside scores like the quick Sequential Organ Failure Assessment (qSOFA) score.

Objectives:

The study aimed at assessing clinical parameters as predictors of in-hospital mortality for people with HIV (PWH) admitted for routine medical care.

Study design:

A prospective observational study of all PWH admitted to medical wards at Kamuzu Central Hospital, Lilongwe, Malawi.

Methods:

WHO danger signs and qSOFA score were determined at the first encounter, CD4 count tests were performed, and discharge outcomes were recorded. The discriminatory power of different scores for predicting in-hospital mortality was assessed using the area under receiver-operating-characteristic curves (AUROCs).

Results:

From November 2022 to May 2023, 401 adults aged ⩾18 years were admitted. Advanced HIV disease (CD4 < 200 cells/mm³) was present in 55.2% (95% CI 50.2–60.2). Overall, in-hospital mortality was 25.7% (95% CI 21.3–30.0). Neither sex, age, CD4 count, nor BMI < 18.5 was significantly associated with mortality. Both the WHO score and qSOFA score were significantly associated with increasing mortality. AUROC for WHO score and qSOFA were 0.68 (95% CI 0.61–0.75) and 0.71 (95% CI 0.64–0.78), respectively. Including BMI or CD4 did not significantly improve AUROC. Using only the individual danger sign “inability-to-walk-unaided” yielded a similar AUROC of 0.68 (95% CI 0.61–0.75).

Conclusion:

Increasing WHO danger sign scores were associated with in-hospital mortality; adding BMI or CD4 did not improve predictive accuracy. Notably, the predictive information derived from a single parameter—inability-to-walk-unaided—was as effective as the complete WHO score and was easier to obtain. Given the challenges in comprehensive vital sign recording, this simple measure may prove valuable in triaging PWH admitted to hospitals in resource-limited settings such as Malawi.

Keywords

in-hospital mortality predictive scores PWH

Introduction

People with advanced HIV admitted to hospitals face a high mortality risk. In South Africa, inpatient mortality among people with HIV (PWH) in 2017 was 20.5%, compared to 10.5% in HIV-seronegative or status-unknown patients. The same study showed that despite overall improving survival over time, PWH mortality remained high in 2020 (17.8% vs. 8.1%).¹ PWH dying in the hospitals are typically younger (median age 44 years) compared to their HIV-negative counterparts (median 64 years).¹ This pattern is also observed across other African settings as demonstrated in Sierra Leone.² Historical data from our own institution in Malawi showed that PWH admitted to medical wards in 2016 had higher in-hospital mortality than HIV-negative patients (adjusted risk ratio 1.6–2.4).³

Tuberculosis is the leading cause of hospitalization and death for PWH in Malawi³ and sub-Saharan Africa,⁴ followed by other infections and opportunistic malignancies related to advanced HIV disease (AHD)^5,6 typically defined as CD4 cell counts below 200 cells/mm.⁷

Preventable causes of in-hospital mortality in sub-Saharan Africa include health system failures such as late presentation of HIV cases and problems associated with loss-to-follow-up after discharge.⁸ Individual hospital factors may include low rates of in-hospital HIV testing, poor laboratory capacity limiting CD4 testing, opportunistic infection diagnostics, and delayed in-hospital initiation of antiretroviral therapy.⁹ To address these gaps, Lighthouse clinics have established a routine service delivery model supporting care for adult PWH admitted to referral hospitals in Malawi that include providing point-of-care HIV and AHD testing as well as management and treatment of key opportunistic infections.¹⁰

To focus care efforts, predictors of poor outcomes among in-patients with HIV are desirable for risk stratification of treatment and diagnostic efforts. Ideally, these would be derived from simple signs or scores routinely elicited at admission. WHO has suggested four “danger signs” to classify patients as seriously ill: respiratory rate > 30/min, heart rate > 120/min, temperature > 39°C, and being unable-to-walk-unaided.¹¹ Other scoring systems like the Universal Vital Assessment (UVA) score^12,13 and the quick Sequential Organ Failure Assessment (qSOFA) score¹⁴ are less HIV-specific but designed to rapidly identify those at high risk of poor outcomes in the context of sepsis. Data suggest qSOFA could aid in risk stratification in resource-limited, HIV-high-prevalence settings.^15–17 The evidence base for these parameters and prediction tools is limited, and their implementation under routine care conditions is rarely assessed. Our study aimed to prospectively evaluate the predictive value of different score parameters for inpatient mortality in PWH admitted to a tertiary care hospital.

Methods

Study design

We conducted a prospective observational study of all adults with HIV consecutively admitted at Kamuzu Central Hospital (KCH) medical wards for a period of 6 months from November 2022 to May 2023, with an aim to assess predictors of hospital mortality under routine care. Data collected from patients followed standard admission forms used for patients with PWH in the medical wards of the hospital. No additional information apart from admission data was obtained. No additional tests were performed deviating from the standard of care.

Study setting

KCH is one of the large tertiary hospitals in Malawi, situated in Lilongwe, the capital city of Malawi. It is a referral site for nine district hospitals within the central region. The hospital is a hub for all specialized services and comprise medical, pediatric, and surgical wards as well as obstetrics and gynecology units. It is a public facility financed by the government of Malawi, with some departments supported by bilateral donors. The medical wards admit up to 6700 patients per year, with 30%–40% of those admitted having HIV infection. Lighthouse Trust is a local organization mandated to provide comprehensive, specialized HIV integrated services and operates centers of excellence (CoEs), HIV clinics within tertiary hospital premises, managing complex cases. Malawi has 991,600 PWH on ART majority of which are served in primary health care facilities called health centers. In the event of a need for further review and assessment, they are referred to district hospitals (secondary level care), complicated cases eventually get referred to central hospitals (tertiary level care) either to the Lighthouse CoEs for advanced diagnostics and review or at the central hospital medical wards if in critical condition requiring admission. The KCH medical ward teams received routine support from the Lighthouse clinic, following a previously described service delivery model that aids the care of HIV-positive patients admitted to referral hospitals.¹⁰

Study participants evaluation

All admitted PWH underwent an assessment including CD4 counts using the Alere PIMA™ (Abbott, 100 Abbott Park Road Abbott Park, IL 60064, USA) point-of-care analyzer as part of standard care practice. For those with CD4 counts < 200 cells/mm³ (AHD), urine-lipoarabinomannan (LAM); Alere, Abbott, USA, and serum-Cryptococcal Antigen (CrAg) (IMMY, 2005 North 10th Street Norman, OK 73071, USA) tests were conducted. Body mass index (BMI), oxygen saturation (SpO₂), WHO danger signs, and qSOFA criteria were recorded at the first encounter. The WHO danger signs used are (a) respiratory rate higher than 30/min (RR > 30), (b) heart rate above 120/min (HR > 120), (c) body temperature above 39°C (T > 39) determined with a digital infrared forehead thermometer and (d) being “unable-to-walk-unaided” as observed by the clinician. qSOFA criteria used are (a) systolic blood pressure below 100 mmHg (BP < 100) measured with an electronic oscillatory device, (b) respiratory rate higher than 22 breaths/min (RR > 22/min), and (c) Glasgow Coma Scale below 15 points (GCS < 15). Patients received routine care as determined by the medical team. Treatments for opportunistic infections and antiretroviral therapy were administered according to Malawi Clinical Guidelines for the management of HIV in adults and children.¹⁸ Patient outcomes during the hospital stay were recorded to determine mortality.

Data analysis

All patients admitted and seen in this period were enrolled in the study. Data were collected on patient assessment forms (Supplemental Material 1) and anonymously extracted into a database. We analyzed baseline characteristics using descriptive statistics. Numeric variables were expressed as median with interquartile ranges (IQR) or mean with 95% confidence intervals (95% CI) and compared using the Mann–Whitney U-test. Categorical variables were expressed as proportions with Wald 95% CI and compared using the Chi-square test. The area under receiver-operating-characteristic curves (AUROCs) with 95% CI was used to assess the discrimination of the different scores for in-hospital mortality. Data were analyzed using SPSS software version 29.0 (IBM Inc., Chicago, IL, USA). A p-value less than 0.05 was considered statistically significant sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) for the single finding “inability-to-walk-unaided” for in-hospital mortality were calculated using the standard formula; standard logit confidence intervals are reported.¹⁹

Results

During the 6 months study period, 401 PWH aged ⩾18 years were admitted to the medical ward and assessed by the team. Demographic information is shown in Table 1. AHD was detected in 55.2% (95% CI 50.2–60.2) of patients. The prevalence of positive urine-LAM and serum-CrAg is shown in Table 1. Overall, 103 patients died, resulting in an in-hospital mortality was 25.7% (95% CI 21.3–30.0). Many deaths (65.0% (95% CI 55.0–74.1)) occurred within the first week of hospitalization. In univariate analysis, neither sex, age, nor BMI < 18.5 was statistically significantly associated with mortality.

Table 1.

Characteristics of PWH admitted to medical wards at Kamuzu Central Hospital and followed till the end of hospital stay from November 2022 to March 2023.

Patient characteristics*	Overall (n = 401)	Died in hospital (n = 103)	Discharged alive (n = 298)	p
Gender (female/male)	201/200 (50.1%/49.9%)	49/54 (47.6%/52.4%)	152/146 (51.0%/49.0%)	0.54
Age (years, median, IQR)	39 (30–49)	41 (34–50)	38 (29–48)	0.10
BMI (kg/m², median, IQR)	21.5 (19.1–23.8)	22.0 (19.1–25.2)	21.4 (19.2–23.4)	0.15
SpO₂ (%, median, IQR)	98 (96–99)	98 (90–99)	98 (97–99)	0.06
CD4 (cells/mm³, median, IQR)	167 (76–338)	132 (54–279)	181 (83–354)	0.05
CD4 < 200 cells/mm³ (%, 95% CI)	55.2 (50.2–60.2)	61.8 (51.4–71.2)	52.9 (47.0–58.7)	0.13
Hospital stay (days, median, IQR)	6 (3–10)	4 (3–9)	7 (4–10)	<0.01
Tests in patients with CD4 < 200
Serum-CrAg + (%, 95% CI)	10.6 (6.8–15.5)	13.7 (6.1–25.3)	9.5 (5.4–15.2)	0.45
Urine-LAM + (%, 95% CI)	28.0 (22.0–34.6)	42.8 (29.7–56.7)	22.5 (16.1–30.0)	<0.01
WHO danger signs
Unable to walk	212/401 (52.9%)	81/103 (78.6%)	131/298 (44.0%)	<0.01
Respiratory rate > 30/min	3/375 (0.7%)	2/97 (1.9%)	1/278 (0.3%)	0.11
Heart rate > 120/min	18/395 (4.5%)	7/99 (6.8%)	11/296 (3.7%)	0.17
Temp > 39°C	3/388 (0.7%)	1/98 (1.0%)	2/290 (0.7%)	0.75
qSOFA criteria
Respiratory rate > 22/min	83/375 (22.1%)	33/97 (32.0%)	50/278 (16.8%)	<0.01
Blood pressure < 100 mmHg	86/399 (21.6%)	27/103 (26.2%)	59/296 (19.9%)	0.18
Glasgow Coma Scale < 15	118/378 (31.2%)	56/96 (58.3%)	62/282 (21.9%)	<0.01

Qualitative variables were compared between deceased and alive patients using Chi²-test; for quantitative variables, the Mann–Whitney-U-test was applied.

IQR, interquartile ranges; LAM, lipoarabinomannan; PWH, people with HIV; qSOFA, quick Sequential Organ Failure Assessment.

Inpatient mortality stratified by CD4 count was 28.9% (95% CI 22.8–35.0) for patients with CD4 < 200 and 22.0% (95% CI 15.8–28.3) in patients with CD4 > 200; this difference was not statistically significant (p = 0.13). Among patients with CD4 < 200, mortality was significantly higher in urine-LAM positive patients (41.4% (95% CI 29.6–54.2)) compared to urine-LAM negative patients (21.5% (95% CI 15.6–28.7), p = 0.005). No statistically significant difference in mortality was observed between serum-CrAg positive (34.8% (95% CI 18.8–55.1)) and serum-CrAg negative (25.9% (95% CI 20.2–32.5), p = 0.36) patients in the AHD group.

Clinical findings for the WHO score and the qSOFA score at the first encounter are shown in Table 1. The complete WHO danger score was available for 359 PWH: 189 (52.6%) had one WHO danger sign, 9 (2.5%) had two, and none had more. The complete qSOFA score was calculated for 351 PWH: 160 (39.9%) had a score of zero, 130 (32.4%) of 1, 54 (13.5%) of 2, and 7 (1.7%) of 3. Table 2 shows in-hospital mortality stratified by scores, with higher scores significantly associated with increasing mortality.

Table 2.

In-hospital mortality of PWH in relation to WHO danger and qSOFA score.

Score	WHO danger score (n = 359)	qSOFA criteria (n = 351)
Score 0	11.8% (7.7–17.7)	12.5% (8.2–18.5)
Score 1	34.4% (28.0–41.4)	30.0% (22.7–38.4)
Score 2	66.6% (35.4–87.9)	50.0% (37.1–62.9)
Score 3	—	57.1% (25.0–84.2)

Trend for both the WHO danger score and qSOFA criteria: p < 0.001.

PWH, people with HIV; qSOFA, quick Sequential Organ Failure Assessment.

Receiver-operator curves for the WHO danger score and the qSOFA score are shown in Figure 1, with nearly identical AUROCs. To explore whether diagnostic performance could be improved, other parameters (CD4 < 200, CD4 < 100, BMI < 18.5 kg/m²) were added to the WHO score without meaningfully improving the AUROC. SpO₂ was excluded due to missing data in 191/401 patients. Using only the individual danger sign “inability-to-walk-unaided” yielded an AUROC of 0.68 (95% CI 0.61–0.75), similar to the complete WHO danger and qSOFA scores.

Figure 1.

Receiver-operator curves for qSOFA score, WHO danger score, expanded WHO danger scores, and inability-to-walk only.

The single finding “inability-to-walk-unaided,” which was elicited for all patients, predicted in-hospital death with a sensitivity of 78.6% (95% CI 69.4–86.1) and a specificity of 56.0% (95% CI 50.2–61.8). Given the 25.7% mortality seen in our cohort, this parameter had a negative predictive value (NPV) of 88.3% (95% CI 83.8–91.8) and a positive predictive value (PPV) of 38.2% (95% CI 34.4–42.1).

Discussion

Our results reveal a prevalence of AHD (54.4%) among PWH admitted to a Malawian referral hospital. This proportion significantly exceeds the national average in the Malawi National ART cohort, which ranged from 28% and 33% in 2022 and 2023, respectively, based on program performance reports. The 25% mortality rate among admitted PWH, with many deaths occurring early during hospitalization, underscores the urgent need to optimize in-hospital care for this vulnerable population.^4,6,8

In our study, the number of WHO danger signs present at first encounter was associated with in-hospital mortality. Although originally designed to stratify PWH for TB diagnostic algorithms,¹⁰ a higher “WHO danger score” was associated with increased mortality under routine care conditions. This aligns with findings from South African studies, where the AUROC for the WHO danger score was 0.65 (17), similar to our observations. Adding parameters such as BMI or CD4 counts marginally increased the AUROC to 0.70, unlikely to substantially improve predictive performance. Similarly, using the additional information on BMI, CD4 count, hypotension, confusion, and ART status augmented the AUROC to 0.75 in the South African analysis.¹⁷ However, an inclusion criterion for the South African study was the presence of at least one WHO danger sign, so despite the similar findings, the data may not be fully comparable.

A higher qSOFA score was also significantly associated with increasing mortality, consistent with previously reported findings from our institution (15) and in Tanzania (16) and Rwanda (17).¹⁴ The UVA score outperformed qSOFA (AUROC 0.77 vs. 0.69 (12) and 0.66 (13) respectively) but requires oxygen saturation data, which was missing for 47.6% (191/401) of our patients due to equipment unavailability. The HIV In-hospital Mortality Prediction (HIV-IMP) risk score,^20,21 developed in South Africa, showed good discrimination (AUROC 0.83, 95% CI 0.78–0.88), but requires in addition to oxygen saturation, laboratory values such as creatinine, lactate, and albumin which is often unavailable in resource-limited settings.

Our assessment revealed that even simple vital signs were not consistently recorded, likely due to time constraints and equipment shortages. This challenges the implementation of robust screening protocols. We found that a single WHO danger sign—“inability-to-walk-unaided”—was as predictive as the complete WHO danger score while being easier to obtain. This parameter could enable risk identification by lay health workers, such as HIV diagnostic assistants, who are widely involved in our setting to perform HIV testing and status ascertainment.²² In-patients who are “unable-to-walk-unaided” could immediately be reported to nurses and clinicians and could thus be prioritized for AHD diagnostic tests and assessments.

Limitations

This evaluation was done in one of the four tertiary hospitals in Malawi, and results may not be easily generalizable to district hospitals and other health facilities, as patients with more severe disease may be over-represented. No sample size was predefined; it was conveniently determined by patients admitted. At the end of the 6 months period, the observation was ended and available data was analyzed. As missing data was considered a reflection of routine conditions and feasibility in the setting, no statistical sensitivity analysis was performed. Reasons for patient admissions were recorded in the patients’ files but not routinely extracted by the HIV team; current ART, adherence, and recent viral load results were commonly not available. Therefore, the impact of underlying conditions, affected organ systems, and previous ART on mortality cannot be discerned; their effect would be worthwhile to assess in future studies.

Conclusion

Increasing WHO danger sign scores are associated with in-hospital mortality, with BMI or CD4 count additions not improving predictive power. The easily obtainable “inability-to-walk-unaided” parameter performed comparably to the complete WHO danger score. Given the challenges in comprehensive vital sign recording, this simple measure may prove valuable in triaging PWH admitted to hospitals in resource-limited settings such as Malawi.

Supplemental Material

sj-docx-1-tai-10.1177_20499361251341385 – Supplemental material for Inability-to-walk-unaided—a single WHO danger sign predicts in-hospital mortality in people with HIV under routine care conditions in a low-resource setting

Supplemental material, sj-docx-1-tai-10.1177_20499361251341385 for Inability-to-walk-unaided—a single WHO danger sign predicts in-hospital mortality in people with HIV under routine care conditions in a low-resource setting by Ethel Rambiki, Agness Thawani, Davis Kapenga, Chikaiko Malunda, Boniface Mseke, Patrick Mpesi, Prakash Ganesh, Hans-Michael Steffen, Tom Heller and Claudia Wallrauch in Therapeutic Advances in Infectious Disease

Footnotes

Acknowledgements

Special thanks to Professor Kwasi Torpey for his technical review of the manuscript and his comments that strengthened the article. Additionally, warm gratitude to Dr Jonathan Ngoma, who was serving as Kamuzu Central Hospital Director General at the time of this evaluation, and the Lighthouse Room 5 team, as well as we Ministry of Health inpatient team for their endless dedication.

Author’s note

Data from this report was presented as an oral presentation at the INTEREST conference, Cotonou, Benin, 2024.

Declarations

ORCID iDs

Ethel Rambiki

Agness Thawani

Hans-Michael Steffen

Supplemental material

Supplemental material for this article is available online.

References

Lamprecht

Martinson

Variava

Effect of HIV on mortality among hospitalised patients in South Africa. Southern African J HIV Med 2023; 24(1): 9.

Lakoh

Jiba

Kanu

, et al. Causes of hospitalization and predictors of HIV-associated mortality at the main referral hospital in Sierra Leone: a prospective study. BMC Public Health 2019; 19(1): 1320.

Matoga

Rosenberg

Stanley

, et al. Inpatient mortality rates during an era of increased access to HIV testing and ART: a prospective observational study in Lilongwe, Malawi. PLOS One 2018; 13(2): e0191944.

Kayambankadzanja

Schell

Namboya

, et al. The prevalence and outcomes of sepsis in adult patients in two Hospitals in Malawi. Am J Trop Med Hygiene 2020; 102(4): 896–901.

Ghislain

Mushebenge

GAA

Magula

Cause of hospitalization and death in the antiretroviral era in Sub-Saharan Africa published 2008–2018: a systematic review. Medicine (Baltimore) 2021; 100(43): e27342.

Agaba

Digin

Makai

, et al. Clinical characteristics and predictors of mortality in hospitalized HIV-infected Nigerians. J Infect Dev Ctries 2011; 5(5): 377–382.

World Health Organisation (WHO). Advanced HIV disease (AHD), https://www.who.int/teams/global-hiv-hepatitis-and-stis-programmes/hiv/treatment/advanced-hiv-disease (2023, accessed 31 Jan 2025).

Owachi

Akatukunda

Nanyanzi

, et al. Mortality and associated factors among people living with HIV admitted at a tertiary-care hospital in Uganda: a cross-sectional study. BMC Infect Dis 2024; 24(1): 239.

Wajanga

Webster

Peck

, et al. Inpatient mortality of HIV-infected adults in sub-Saharan Africa and possible interventions: a mixed methods review. BMC Health Serv Res 2014; 14(1): 627.

10.

Heller

Damba

Kumwenda

, et al. Implementing advanced HIV disease care for inpatients in a Referral Hospital in Malawi - demand, results and cost implications. Ann Glob Health 2022; 88(1):16.

11.

WHO. Improving the diagnosis and treatment of smear-negative pulmonary and extrapulmonary tuberculosis among adults and adolescents: recommendations for HIV-prevalent and resource-constrained settings [Améliorer le diagnostic et le traitement de la tuberculose pulmonaire à frottis négatif ou extrapulmonaire chez l’ adulte et l’ adolescent : recommandations à l’ intention des pays de prévalence du VIH et disposant de ressources limitées] [Internet], https://iris.who.int/handle/10665/69463 (2007, accessed 31 Jan 2025).

12.

Moore

Hazard

Saulters

, et al. Derivation and validation of a universal vital assessment (UVA) score: a tool for predicting mortality in adult hospitalised patients in sub-Saharan Africa. BMJ Glob Health 2017; 2(2): e000344.

13.

Hazard

Bagenda

Patterson

, et al. Performance of the Universal Vital Assessment (UVA) mortality risk score in hospitalized adults with infection in Rwanda: a retrospective external validation study. PLoS One 2022; 17(3): e0265713.

14.

Rudd

Seymour

Aluisio

, et al. Association of the quick sequential (Sepsis-Related) organ failure assessment (qSOFA) score with excess hospital mortality in adults with suspected infection in low- and middle-income countries. JAMA 2018; 319(21): 2202–2211.

15.

Huson

MAM

Katete

Chunda

, et al. Application of the qSOFA score to predict mortality in patients with suspected infection in a resource-limited setting in Malawi. Infection 2017; 45(6): 893–896.

16.

Boillat-Blanco

Mbarack

Samaka

, et al. Prognostic value of quickSOFA as a predictor of 28-day mortality among febrile adult patients presenting to emergency departments in Dar es Salaam, Tanzania. PLoS One 2018; 13(6): e0197982.

17.

Aluisio

Garbern

Wiskel

, et al. Mortality outcomes based on ED qSOFA score and HIV status in a developing low income country. Am J Emerg Med 2018; 36(11): 2010–2019.

18.

Ministry of Health. Department of HIV, STIs & Viral Hepatitis. Malawi integrated clinical HIV guidelines 5th Edition (2022) - Malawi Integrated Clinical HIV Guidelines 1st Edition (2022) - The Document Management System, https://dms.hiv.health.gov.mw/dataset/malawi-intergrated-clinical-hiv-guidelines-1st-edition-2022/resource/cf0aea4d-1100-4cfa-8588-db9876cece4c (2022, accessed 7 Jan 2025).

19.

Schoonjans

MedCalc, MedCalc’s diagnostic test evaluation calculator, https://www.medcalc.org/calc/diagnostic_test.php (2020, accessed 31 Jan 2025).

20.

Laher

Paruk

Venter

, et al. Predictors of in-hospital mortality among HIV-positive patients presenting with an acute illness to the emergency department. HIV Med 2021; 22(7): 557–566.

21.

Laher

Paruk

Venter

WDF

, et al. Development and internal validation of the HIV In-hospital Mortality Prediction (HIV-IMP) risk score. HIV Med 2022; 23(1): 80–89.

22.

Flick

Simon

Nyirenda

, et al. The HIV diagnostic assistant: early findings from a novel HIV testing cadre in Malawi. AIDS 2019; 33(7): 1215–1224.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.09 MB