Cost-effectiveness of intensive care for hospitalized COVID-19 patients: experience from South Africa.

BACKGROUND: Given projected shortages of critical care capacity in public hospitals during the COVID-19 pandemic, the South African government embarked on an initiative to purchase this capacity from private hospitals. In order to inform purchasing decisions, we assessed the cost-effectiveness of intensive care management for admitted COVID-19 patients across the public and private health systems in South Africa. METHODS: Using a modelling framework and health system perspective, costs and health outcomes of inpatient management of severe and critical COVID-19 patients in (1) general ward and intensive care (GW + ICU) versus (2) general ward only (GW) were assessed. Disability adjusted life years (DALYs) were evaluated and the cost per admission in public and private sectors was determined. The model made use of four variables: mortality rates, utilisation of inpatient days for each management approach, disability weights associated with severity of disease, and the unit cost per general ward day and per ICU day in public and private hospitals. Unit costs were multiplied by utilisation estimates to determine the cost per admission. DALYs were calculated as the sum of years of life lost (YLL) and years lived with disability (YLD). An incremental cost-effectiveness ratio (ICER) - representing difference in costs and health outcomes of the two management strategies - was compared to a cost-effectiveness threshold to determine the value for money of expansion in ICU services during COVID-19 surges. RESULTS: A cost per admission of ZAR 75,127 was estimated for inpatient management of severe and critical COVID-19 patients in GW as opposed to ZAR 103,030 in GW + ICU. DALYs were 1.48 and 1.10 in GW versus GW + ICU, respectively. The ratio of difference in costs and health outcomes between the two management strategies produced an ICER of ZAR 73,091 per DALY averted, a value above the cost-effectiveness threshold of ZAR 38,465. CONCLUSIONS: Results indicated that purchasing ICU capacity from the private sector during COVID-19 surges may not be a cost-effective investment. The 'real time', rapid, pragmatic, and transparent nature of this analysis demonstrates an approach for evidence generation for decision making relating to the COVID-19 pandemic response and South Africa's wider priority setting agenda.

Managing the healthcare demand-supply gap during and after COVID-19: The need to review the approach to healthcare priority-setting in South Africa.

Healthcare demands are rising globally, and regardless of the approach to financing and delivering healthcare services, no country can meet all the healthcare demands of its population. The demand-supply gap for healthcare services in South Africa (SA) is large, particularly for the public sector. The objectives of this article are to examine some of the underlying factors contributing to this gap, and how the COVID- 19 pandemic is likely to impact on them, and to describe why SA needs to adopt an explicit and equity-informed approach to healthcare priority-setting to assist in managing the gap.

Socioeconomic Status and Barriers to the Use of Free Antiretroviral Treatment for HIV/AIDS in Enugu State; South-Eastern Nigeria

With about 3 million people living with HIV; Nigeria has approximately 8of the global burden of HIV cases. In 2009; only about 34of those in need of antiretroviral treatment (ART) were able to access care; which means that Nigeria was far from achieving the United Nations target of `universal access' by 2010. This study aimed to describe the barriers to accessibility and the coping strategies employed to overcome these barriers among users of free ART services overall and by socioeconomic status (SES). Data were collected from 240 people receiving ART at one urban and one peri-urban health facility in Enugu State; south-eastern Nigeria. Information on SES; demographic characteristics; and barriers and coping mechanisms for accessing ART were elicited from the respondents. The high cost of transportation; HIV stigma; and long waiting hours were found to be key barriers to the use of ART services. On average; ART clients spent just under four hours at the clinic during their monthly appointments. The use of personal savings and financial support from relatives were the main means to access treatment. When the data were analysed according to clients' SES; transportation costs were a chief concern among the poorest while those who were better off were more likely to be concerned about stigma and discrimination. These findings should be borne in mind when designing and locating services to maximise ongoing accessibility to ART

Extremely prolonged continuous intraperitoneal infusion of cytosine arabinoside.

Intraperitoneal administration of ara-C produces a peritoneal/plasma concentration ratio of 330-1,000: In principle, optimal tumor-cell kill should be obtained when high ara-C concentrations ar maintained in the environment of the tumor for very long periods of time. A phase 1 study was undertaken to determine the maximum tolerated dose of ara-C that could be given as a continuous i.p. infusion for 3 weeks. A total of 14 patients with refractory malignancies were given 28 courses in the outpatient setting. Ara-C infusions were given using a portable programmable pump (Pancreatec Provider Model 2000). No significant side effects were observed in patients receiving 30 mg/m2 per day (five courses) or 40 mg/m2 per day x 21 days (seven courses). However, at a dose of 60 mg/m2 per day, although 10/16 courses were tolerated for at least 1 week, only 3/16 attempted courses could be continued for the full 3 weeks. The dose-limiting toxicity was chemical peritonitis, which occurred during 7/16 courses at this dose level and required termination of therapy in 4 courses. Myelosuppression was also observed at this dose. There was a large variation in the ara-C and ara-U peritoneal concentrations both within and between patients. The mean peritoneal ara-C concentration increased nonlinearly with ara-C dose whereas the mean ara-U concentration decreased. This study establishes the feasibility and safety of giving a cell-cycle-specific drug intraperitoneally over an extremely prolonged period. For subsequent studies a dose of 40 mg/m2 per day for 21 days is recommended.

Effect of sodium thiosulfate on the pharmacokinetics and toxicity of cisplatin.

Concurrent administration of sodium thiosulfate (STS) can protect against the nephrotoxic effects of even very-high-dose cisplatin (CDDP) (i.e., 270 mg/m2 given intraperitoneally). The effect of STS on the pharmacology and toxicity of CDDP was investigated in patients receiving at each treatment 90 mg of CDDP/m2 intraperitoneally, with STS given concurrently on alternate cycles by the intravenous route. The patients received a total of 38 courses of therapy, 21 without STS and 17 with STS. STS reduced the total exposure to diethyldithiocarbamate-reactive CDDP for the peritoneal cavity and plasma by 36% and 25%, respectively. When given alone, CDDP caused a statistically significant acute reduction in creatinine clearance levels; this reduction was less evident when STS was given. We conclude that, whereas STS does reduce systemic exposure, the magnitude of this effect was not sufficient to account for the ability of STS to protect against high-dose CDDP.

Selective intraperitoneal biochemical modulation of methotrexate by dipyridamole.

Dipyridamole increases the toxicity of methotrexate in a concentration-dependent manner. We hypothesized that concurrent intraperitoneal administration of both drugs would result in high peritoneal concentrations with much lower plasma concentrations, permitting a selective increase in the activity of methotrexate against intraperitoneal tumors without enhancing systemic toxicity. Initially, 2.16 mg/m2/d methotrexate and 12 mg/m2/d dipyridamole were delivered together as a constant intraperitoneal infusion for 48 hours. With escalation of chemotherapy, eventually 4.32 mg/m2/d methotrexate was administered for 168 hours. Forty-seven courses were administered to 18 patients. The mean peritoneal to plasma concentration ratios of methotrexate and non-protein bound dipyridamole were 71.6 +/- 34.8 and over 2,300, respectively. Chemical peritonitis was the dose-limiting toxicity. Three patients had some evidence of a response (two with decreasing tumor markers, and the third with a reduction in ascites). We conclude that the drug concentrations are in an appropriate range for selective intraperitoneal biochemical modulation of methotrexate, and that it is feasible to expose tumors confined to the peritoneal cavity to these drugs for long periods of time.

Phase I/pharmacokinetic study of thioguanine administered as a 48-hour continuous intraperitoneal infusion.

Thioguanine (TG) is an antimetabolite with established antileukemic activity. The most pharmacologically rational manner of TG administration is continuous infusion. Intraperitoneal (IP) delivery of TG provides the opportunity to definitively test the concept of high-dose, long-term antimetabolite administration. The high systemic clearance and in vitro activity of TG against ovarian carcinoma suggested that it would be an excellent candidate for IP administration as a prolonged infusion. TG was administered as a 48-hour continuous IP infusion in this phase I/pharmacokinetic study. TG infusions were administered using a portable, programmable pump (Pancretec Provider Model 2000; Pancretec, Inc, San Diego). Twenty-five patients were treated. At a dose of 900 mg/m2/48 h, TG produced unacceptably severe myelosuppression. The dose-limiting toxicity was granulo-cytopenia. Other toxicities were mild: emesis, alopecia, skin rashes, and photosensitivity reactions. IP TG did not produce chemical peritonitis, hepatotoxicity, or mucositis. The pharmacokinetics of IP TG were determined in 16 patients. TG levels were measured by reverse-phase high-performance liquid chromatography (HPLC). At steady state, the mean peritoneal to plasma TG ratio was 1,800 at the maximum tolerated dose (MTD). Steady-state TG levels in the peritoneal cavity and plasma were 2 mmol/L and 1.1 mumol/L, respectively, at a dose of 744 mg/m2. The elimination half-life of TG from the peritoneal cavity was one hour. TG exhibited linear pharmacokinetics over the dosage range investigated. Encouraging clinical activity was seen with IP TG. There was one partial response (PR) and four minor responses (MR). TG can be safely administered by the IP route. The recommended dose for phase II testing is 744 mg/m2/48 h. IP TG has a favorable pharmacokinetic advantage and has demonstrated encouraging clinical activity. Further studies of IP TG infusions are warranted.

Phase I/pharmacokinetic study of intraperitoneal cisplatin and etoposide.

We administered cisplatin and etoposide by peritoneal dialysis to 39 patients with i.p. malignancies in order to investigate the toxicity, pharmacokinetics, and clinical activity of this 2-drug combination. All patients received i.v. sodium thiosulfate concurrently with the i.p. chemotherapy. Myelosuppression, nausea, vomiting, and malaise were the primary toxicities encountered. The maximum tolerated dose of etoposide was 350 mg/m2, when administered with a fixed dose of cisplatin, 200 mg/m2. Although the total (free and protein-bound) etoposide exposure for the peritoneal cavity was only 1.5-fold greater than that for the plasma, the free (non-protein bound) etoposide peritoneal exposure was 65-fold greater than the plasma. Tumor regressions were noted in patients with ovarian and pancreatic carcinomas. This study is the first demonstration of the large pharmacokinetic advantage that exists for the i.p. administration of highly protein-bound drugs, and it also documents the clinical activity of i.p. cisplatin and etoposide.