A critical analysis of discovery health's claims-based risk adjustment of mortality rates in South African private sector hospitals

In 2019, Discovery Health published a risk adjustment model to determine standardised mortality rates across South African private hospital systems, with the aim of contributing towards quality improvement in the private healthcare sector. However, the model suffers from limitations due to its design and its reliance on administrative data. The publication's aim of facilitating transparency is unfortunately undermined by shortcomings in reporting. When designing a risk prediction model, patient-proximate variables with a sound theoretical or proven association with the outcome of interest should be used. The addition of key condition-specific clinical data points at the time of hospital admission will dramatically improve model performance. Performance could be further improved by using summary risk prediction scores such as the EUROSCORE II for coronary artery bypass graft surgery or the GRACE risk score for acute coronary syndrome. In general, model reporting should conform to published reporting standards, and attempts should be made to test model validity by using sensitivity analyses. In particular, the limitations of machine learning prediction models should be understood, and these models should be appropriately developed, evaluated and reported.

A critical analysis of Discovery Health's claims-based risk adjustment of mortality rates in South African private sector hospitals.

In 2019, Discovery Health published a risk adjustment model to determine standardised mortality rates across South African private hospital systems, with the aim of contributing towards quality improvement in the private healthcare sector. However, the model suffers from limitations due to its design and its reliance on administrative data. The publication's aim of facilitating transparency is unfortunately undermined by shortcomings in reporting. When designing a risk prediction model, patient-proximate variables with a sound theoretical or proven association with the outcome of interest should be used. The addition of key condition-specific clinical data points at the time of hospital admission will dramatically improve model performance. Performance could be further improved by using summary risk prediction scores such as the EUROSCORE II for coronary artery bypass graft surgery or the GRACE risk score for acute coronary syndrome. In general, model reporting should conform to published reporting standards, and attempts should be made to test model validity by using sensitivity analyses. In particular, the limitations of machine learning prediction models should be understood, and these models should be appropriately developed, evaluated and reported.

Implementing a multi-faceted framework for proprietorship of hand hygiene compliance in a network of South African hospitals: leveraging the Ubuntu philosophy.

BACKGROUND: Given the lack of hospital-wide ownership and shortage of nurses, the ideal model for large-scale implementation of hand hygiene (HH) behaviour change in low- and middle-income countries is unknown. AIM: The aim of the multi-modal strategy was to engender hospital accountability for HH compliance. METHODS: The quasi-experimental study was conducted in 50 South African hospitals (November 2015 to July 2017) and involved five overlapping phases: executive governance and corporate behaviour change; group-wide systematic situational analysis; development of an electronic-assisted direct-observed data collection and analysis application; launch and implementation; and accountable governance. Measurement of intra- and inter-hospital variance to six HH opportunities was calculated and data compliance dashboards were e-mailed weekly to hospital leadership teams to provide feedback of recorded HH compliance and behaviour to frontline teams. Baseline comparison (July 2016) of compliance was compared versus post-implementation (July 2017). FINDINGS: Baseline HH compliance of ≤60% was documented for 16% (8/50) of hospitals, whereas overall, 48% (24/50) of hospitals demonstrated a significant improvement (P < 0.01). Over the 13-month observation period, 523,422 observations were recorded with a mean rate of 277 ± 223 observations per 1000 patient-days. The group mean composite compliance improved by 7.8% (P < 0.01) from 77.4% ± 12.8 to 85.2% ± 8.8 between July 2016 and July 2017, respectively. CONCLUSION: Implementation of a multi-faceted HH model in a large, diverse group of South African hospitals translated into changes in the organizational systems and accountability, standardized HH compliance management and feedback that led to HH proprietorship.

Impact of iron loading on the activity of isoniazid or ethambutol in the treatment of murine tuberculosis.

OBJECTIVE: To assess the impact of iron loading on the activity of isoniazid and ethambutol in the treatment of murine tuberculosis. DESIGN: Iron-loaded and iron-normal female Balb/C mice infected with 1.5 x 10(7) colony forming units of Mycobacterium tuberculosis were treated with either isoniazid or ethambutol for 28 days. RESULTS: For both treatments, the outcome was impaired by the iron loading: bactericidal activity of isoniazid was partially but significantly reduced and ethambutol bactericidal activity was totally inhibited. CONCLUSION: The treatment of tuberculosis in patients with iron loading should be longer than for normal patients or should contain an additional drug.

Epidemiology of invasive aspergillosis in France: a six-year multicentric survey in the Greater Paris area.

Invasive aspergillosis is the most prevalent mould infection. An epidemiological surveillance network was set up in 18 teaching hospitals in Paris and the Greater Paris area. Prospective surveillance was conducted between 1994 and 1999. Between 1994 and 1997 cases were categorized as proven or probable aspergillosis and then the European Organization for Research and Treatment of Cancer/Mycoses Study Group criteria were used. The authors analysed 621 cases (115 proven, 506 probable). No seasonal variation was found. Haematological disorders (73%) including stem-cell transplantation (36%), solid-organ transplantations (10%) and AIDS (9%) were the main underlying conditions. The crude mortality was 63%. Incidence of IA was 8% (CI(95): 6.5-9.5) in acute myelocytic leukaemia and 6.3% (CI(95): 4.3-8.3) in acute lymphocytic leukaemia. Incidence was 12.8% (CI(95): 10.8-14.8) following allogeneic stem-cell transplantation and 1.1% (CI(95): 0.7-1.5) following autologous stem-cell transplantation. In solid-organ recipients incidence ranged from 11% following heart-lung transplantation and small bowel to 0.4% following kidney transplantation. Incidence in HIV infected patients ranged from 0.02 to 0.13% per annum. This large series confirmed that patients with haematologic disorders and transplantations are the most at risk for IA.

Trial of oral fumagillin for the treatment of intestinal microsporidiosis in patients with HIV infection. ANRS 054 Study Group. Agence Nationale de Recherche sur le SIDA.

OBJECTIVE: Intestinal microsporidiosis caused by Enterocytozoon bieneusi is a cause of chronic diarrhoea in patients with HIV infection for which there is no current therapy. This study was designed to assess the safety and efficacy of oral fumagillin in this infection. DESIGN: A dose-escalation trial. METHODS: Twenty-nine HIV-infected patients with E. bieneusi infection were consecutively enrolled in the trial. Oral doses of fumagillin were given to four groups of patients for 14 days: 10 mg/day (group 1), 20 mg/day (group 2), 40 mg/day (group 3), and 60 mg/day (group 4). Patients were seen at weeks 1, 2, 4 and 6 to assess safety and efficacy. Efficacy was assessed primarily by the clearance of microsporidia from stools and follow-up duodenal biopsies. RESULTS: Thirteen patients complained of abdominal cramps, vomiting or diarrhoea during the study, and three patients had fumagillin withdrawn because of adverse events. Thrombocytopenia, neutropenia and hyperlipasaemia were the most frequent biological adverse events. Twenty-one out of 29 patients transiently cleared microsporidia from their stools during the study. By week 6, however, all patients in groups 1, 2 and 3 had parasitic relapse. Interestingly, eight out of 11 (72%) patients treated with 60 mg/day (group 4) apparently cleared microsporidia from their gastrointestinal tract and gained weight. No parasitic relapse was documented in these eight patients during a mean follow-up of 11.5 months. CONCLUSION: Treatment with fumagillin at 60 mg/day for 14 days has promise as an effective oral treatment for E. bieneusi infections.

[Piercing and its infectious complications. A public health issue in France]. / Le piercing et ses complications infectieuses. Un enjeu de santé publique en France.

CHARACTERISTIC FEATURES: Piercing, an act that modifies the body, has progressed considerably in France over the last few years. The population involved has grown and become more diversified. Performed with a solid needle or a catheter, a wide variety of anatomic localizations are concerned, particularly the nose, ears, and navel. The shape of the "rings", generally made of surgical steel, niobium or titanium, varies greatly. Wound healing by epithelialisation can take up to several months. INFECTIOUS RISK: Between 10% and 20% of all piercings lead to a local infection. The most commonly found causal agests are Staphylococcus aureus, group A Streptococcus and Pseudomonas sp. These germs can cause severe life-threatening complications even in common localizations (earlobe). Viral transmission is another risk (hepatitis B, hepatitis C, hepatitis delta, HIV). A few cases of fatal fulminant hepatitis have been described immediately after piercing. SAFETY MEASURES: Generally performed under less than desirable sanitary conditions, safety measures are needed for piercing. Among professional "piercers", a certain number have emphasized the need for providing their clients with safer services. The prevention of infection risk should be a priority for all. Work along this line has been done in the United States and Canada. In light of the impact on public health, it is important to rapidly develop guidelines and regulations for piercing in France. Both professional piercers and health care workers should participate in developing these safety measures in order to assure their implementation.

Long-term central venous catheter infection in HIV-infected and cancer patients: a multicenter cohort study.

OBJECTIVES: To evaluate and compare the risk of long-term central venous catheter (CVC) infection in human immunodeficiency virus (HIV)-infected and cancer patients. DESIGN: Prospective multicenter cohort study based on active surveillance of long-term CVC manipulations and patient outcome over a 6-month period. SETTING: Services of infectious diseases and oncology of 12 university hospitals in Paris, France. PARTICIPANTS: In 1995, all HIV and cancer patients with solid malignancy were included at the time of long-term CVC implantation. RESULTS: Overall, 31.6% of long-term CVC infections were identified in 32% of 201 HIV and 5% of 255 cancer patients. Most were associated with bacteremia, most commonly coagulase-negative staphylococci. The long-term CVC time-related infection risk was greater in HIV than in cancer patients (3.78 vs 0.39 infections per 1,000 long-term CVC days; P<.001). The independent risk factors of long-term CVC infection were as follows: in HIV patients, frequency of long-term CVC handling and neutropenia; in cancer patients, poor Karnofsky performance status; in both HIV and cancer patients, recent history of bacterial infection. The risk of long-term CVC infection was similar for tunneled catheters and venous access ports in each population. CONCLUSIONS: Prevention of long-term CVC infection should focus first on better sterile precautions while handling long-term CVC, especially in HIV patients who have frequent and daily use of the long-term CVC.

Acute renal failure in the course of HIV infection: a single-institution retrospective study of ninety-two patients and sixty renal biopsies.

BACKGROUND: Acute renal failure syndromes are frequently encountered in patients with human immunodeficiency virus (HIV) infection. Most reported cases of acute renal failure are related to acute tubular necrosis, but many other causes of renal failure have been described in these patients. METHODS: The present work is a single-institution retrospective study of 92 HIV-infected patients with acute or rapidly progressing renal failure. In 60 cases, a renal biopsy was performed. For each patient we analysed clinical and pathological data, as well as the short-term prognosis. RESULTS: Ten different causes of acute or rapidly progressing renal failure were documented: (i) haemolytic uraemic syndrome (32 patients); (ii) acute tubular necrosis either of ischaemic-toxic origin (18 patients) or due to rhabdomyolysis (six patients); (iii) obstructive renal failure which was either extrinsic (two patients), drug-induced (13 patients) or secondary to paraprotein precipitation (one patient); (iv) HIV-associated nephropathy (14 patients); (v) acute interstitial nephritis (two patients); (vi) various glomerulonephritis (four patients). In most cases, renal failure was severe (the mean creatinine clearance at entry was 12 ml/min). Most patients had a significant improvement in renal function with only symptomatic treatment. Eighteen per cent of the patients died within 2 months of the diagnosis of renal failure. Renal biopsy seems important for the diagnosis but also for the prognosis, at least in the cases of haemolytic-uraemic syndrome, HIV-associated nephropathy and drug-induced micro-obstructive renal failure. CONCLUSION: Vascular and glomerular diseases are frequent causes of acute or rapidly progressing renal failure in HIV-infected patients. Renal biopsy appears to be safe and useful for the diagnosis and the prognosis of the renal failure. High mortality rate is only observed in patients with ischaemic/toxic causes of acute renal failure.

Effect of indinavir on HIV-related wasting.

OBJECTIVE: To study the effect of the protease inhibitor indinavir on body weight and body composition of subjects with HIV-related wasting. DESIGN: Prospective measurement of body weight in patients who had wasting and were treated with indinavir. A subgroup of 16 representative patients also underwent a metabolic study that included measurements of body composition (skinfolds and bioelectrical impedance) and food intake. Seven from this subgroup who did not have chronic diarrhoea also underwent indirect calorimetry for measurement of resting energy expenditure; the nine patients with wasting and chronic diarrhoea had measurements of faecal losses and intestinal permeability using the lactulose-mannitol test. SETTING: A tertiary care university hospital. PATIENTS: Two hundred and fourteen HIV-infected patients with wasting (less than 95% of usual body weight) had their body weight measured at day 0; 186 patients had a second body weight measurement within the first 100 days of treatment, and 160 patients were weighed a third time, at a median of 176 days. RESULTS: Body weight increased significantly (P < 0.0001) during treatment, whatever the degree of weight loss at baseline. After a median of 176 days on treatment, body weight had increased in 119 out of the 160 patients followed (74.4%; mean weight gain, 6.3+/-SD 3.8 kg; range, 1-18 kg), had not changed in 13 (8.1%) and had fallen in 28 (17.5%; mean weight loss, 4.2+/-3.0 kg; range, 1-12 kg), relative to baseline. Overall, 119 out of the 214 patients (55.6%) from the initial population gained weight. Fat mass, fat-free mass and body cell mass increased significantly in the 16 patients who underwent metabolic studies, together with energy, protein and lipid intake. In the patients with chronic diarrhoea, intestinal permeability improved but there was no change in intestinal losses. In patients who had wasting but not chronic diarrhoea, resting energy expenditure did not change significantly. Body weight changes correlated with changes in the CD4+ cell count (r = 0.882; P = 0.00001) and, to a lesser extent, with changes in the viral load (r = -0.466; P = 0.047). CONCLUSION: Indinavir significantly improved the nutritional status of these patients with HIV-related wasting.

Factors of weight loss in patients with HIV and chronic diarrhea.

Weight loss is significant in patients with HIV and chronic diarrhea. The aim of our study was to test for the links between weight loss, the level of food intake, and the severity of diarrhea and nutrient malabsorption. One hundred and sixteen patients with HIV and chronic diarrhea underwent a standardized gastrointestinal and nutritional evaluation, which included a questionnaire on diarrhea, a prospective estimation of food intake, a measurement of blood parameters and fecal lipid and nitrogen outputs, a stool examination for bacteria and parasites, and upper and lower digestive tract endoscopy. Diarrhea resulted from an infection by Cryptosporidia, Microsporida, or other pathogens in 22%, 20%, and 13% of the patients, respectively. Diarrhea appeared idiopathic in 45% of the patients. A significant negative correlation existed between the severity of weight loss and the levels of nutrient intake (p < .005), and a significant positive correlation between the severity of weight loss and stool frequency (p < .01). Multiple linear regression identified low caloric intake and high stool frequency as predictive of weight loss. No significant correlation was found between weight loss and the parameters of malabsorption, either by bivariate study or multiple regression. These results suggest that, in patients with HIV and chronic diarrhea, the degree of wasting is significantly related to the levels of dietary intake and the clinical severity of diarrhea, but not to the extent of nutrient malabsorption.

Determination of the inhibitory effect on Toxoplasma growth in the serum of AIDS patients during acute therapy for toxoplasmic encephalitis.

In 26 AIDS patients treated for toxoplasmic encephalitis, the inhibitory effect on Toxoxplasma growth of sequential sera taken before and after initiation of therapy was determined using a culture-based immunoassay and compared with pyrimethamine blood levels. A marked inhibition of Toxoplasma growth was observed 1 day after initiation of therapy with pyrimethamine plus sulfadiazine and was maintained between 67% and 93% throughout the 15-day follow-up period. The degree of inhibition was not correlated to pyrimethamine blood levels and seemed highly potentiated when sulfadiazine rather than macrolides was given in combination with pyrimethamine.

Mycobacterium celatum infection in two HIV-infected patients treated prophylactically with rifabutin.

Mycobacterium celatum is a recently described slow-growing species. It was identified on the basis of genomic sequencing that differentiates three types. The present report describes two cases of Mycobacterium celatum type 1 infection in patients with AIDS. Both patients had CD4+ lymphocyte counts of < 10/mm3, were receiving rifabutin prophylaxis, and had attended the same treatment units. The minimum inhibitory concentration of rifabutin for both strains was 8 mg/l, which may account for the failure of prophylaxis. As all type 1 strains have the same pulsed-field gel electrophoresis pattern, nosocomial transmission or acquisition from a common source could not be ruled out.

In vitro and in vivo activities of moxifloxacin and clinafloxacin against Mycobacterium tuberculosis.

On 10% oleic acid-albumin-dextrose-catalase-enriched 7H11 agar medium, the MIC at which 90% of the isolates are inhibited for 20 strains of Mycobacterium tuberculosis was 0.5 microg of sparfloxacin (SPFX) or moxifloxacin (MXFX) per ml and 1.0 microg of clinafloxacin (CNFX) per ml, indicating that the in vitro activities of SPFX and MXFX were virtually identical and were slightly greater than that of CNFX. However, the in vivo activities of these drugs in a murine tuberculosis model differed considerably. Female Swiss mice were infected intravenously with 6.2 x 10(6) CFU of the H37Rv strain and treated for 4 weeks, beginning the next day after infection, with isoniazid (INH) serving as the positive control. By the criteria of 30-day survival rate, spleen weight, gross lung lesion, and mean number of CFU in the spleen, treatment with CNFX at up to 100 mg/kg of body weight six times weekly displayed no measurable effect against M. tuberculosis, whereas both SPFX and MXFX were effective; administration six times weekly of either of the latter two drugs demonstrated dosage-dependent bactericidal effects, as measured by enumeration of CFU in the spleens, and MXFX appeared more bactericidal than the same dosage of SPFX. Of the three fluoroquinolones, only MXFX at 100 mg/kg six times weekly appeared as bactericidal as INH at 25 mg/kg six times weekly. Thus, MXFX may be an important component of the newer combined regimens for treatment of tuberculosis.

Tissue-type plasminogen activator activity in HIV-associated HUS.

BACKGROUND: In children, haemolytic uraemic syndrome (HUS) is associated with high plasma plasminogen activator inhibitor type 1 (PAI-1), which may contribute to the persistence of renal glomerular and arteriolar thrombi. HUS has been described in HIV-infected patients, but the pathophysiology of HUS in these patients is poorly understood. The aim of the study was to investigate plasma fibrinolytic activity in 18 patients with HIV-associated HUS. METHODS: We measured tissue type plasminogen activator (t-PA) and PAI-1 activities in the plasma of 18 HIV-infected patients with biopsy-proven HUS (HIV+/HUS+) and 48 HIV-infected patients without HUS (HIV+/HUS-). RESULTS: Patients with HUS had a significantly higher serum creatinine, a lower platelet count and an increased incidence of cytomegalovirus (CMV) infection (72% of patients HIV+/HUS+, vs 25% of patients HIV+/HUS-). Unexpectedly, plasma PAI-1 activity was similar in both groups. However, t-PA activity was significantly higher in HUS cases (11.5 vs 4.5 U/ml, P=0.001). Patients with CMV infection, with or without HUS, had significantly increased t-PA level (P=0.01). Multivariate analysis identified high t-PA (RR=9.21) and CMV infection (RR=3.36) as risk factors for HUS. CONCLUSION: This study provides evidence that HIV-infected patients with HUS have high plasma t-PA activity. PAI-1 plasma activity is not significantly increased, as opposed to non-HIV-associated HUS. Thus, in the setting of HIV infection, HUS cannot be attributed to decreased fibrinolytic activity.

Ribavirin therapy for adenovirus pneumonia in an AIDS patient.

We report the effectiveness of ribavirin in an AIDS patient with multinodular pneumonia due to adenovirus. A 38-year-old AIDS patient who experienced multiple opportunistic infections and whose CD4 lymphocyte count was 5/mm3 developed bilateral nodular lung opacities. Lung surgical biopsy yielded necrotizing pneumonitis with characteristic nuclear inclusions and positive immunocytology with adenovirus antibodies. Marked clinical and radiological improvement was obtained after intravenous then oral ribavirin. Ribavirin was discontinued after 40 d because of anemia. Relapse of pneumonia with respiratory distress led to death 8 mo later. This observation illustrates a rarely reported pulmonary opportunistic infection in AIDS and the potential value of ribavirin therapy for adenovirus pneumonia.

Potential efficacy of fumagillin in intestinal microsporidiosis due to Enterocytozoon bieneusi in patients with HIV infection: results of a drug screening study. The French Microsporidiosis Study Group.

OBJECTIVE: Intestinal microsporidiosis due to Enterocytozoon bieneusi is a frequent cause of chronic diarrhoea in patients with HIV infection for which there is no available therapy. This study was designed to search for a drug with activity against this organism. DESIGN: Prospective open-labelled Phase II multicentre study. SETTING: University hospitals. PATIENTS: Sixty HIV-infected men with intestinal E. bieneusi infection. INTERVENTIONS: Ten drug regimens were consecutively tested orally for 3 weeks: albendazole plus metronidazole, sulphadiazine plus pyrimethamine, atovaquone, doxycycline plus nifuroxazide, itraconazole, flubendazole, chloroquine, paromomycin, sparfloxacin and fumagillin. Nine evaluable patients per regimen were required, but each patient could be enrolled up to three times in the study. OUTCOME MEASURE: Efficacy was assessed primarily by the clearance of E. bieneusi from stools and intestinal biopsies. The safety of each regimen was also assessed. RESULTS: Only purified fumagillin was able to clear E. bieneusi from stools as well as intestinal biopsies, whereas all other regimens failed to show antiparasitic efficacy. However, only four patients received fumagillin because of drug-induced thrombocytopenia. The four patients who received fumagillin remained free of E. bieneusi infection after a mean follow-up of 10 months. CONCLUSION: Eradication of E. bieneusi from the intestinal tract of patients with HIV infection and persistent immunosuppression is an achievable goal. Our study allowed the identification of oral fumagillin as a potential treatment for intestinal microsporidiosis due to E. bieneusi.

Clinical and bacteriologic impact of rifabutin prophylaxis for Mycobacterium avium complex infection in patients with human immunodeficiency virus infection.

We conducted a prospective observational study to determine the feasibility and impact of rifabutin prophylaxis (300 mg daily) for human immunodeficiency virus-infected patients whose CD4 cell counts were <100/mm3. Three hundred seventy-one patients (65.2% of all patients with CD4 cell counts of <100/mm3 [mean +/- SD, 30 +/- 25/mm3]) received rifabutin prophylaxis for a mean duration +/- SD of 35.5 +/- 34.2 weeks; 198 patients (mean CD4 cell count +/- SD, 51.6 +/- 32/mm3) did not receive prophylaxis. Rifabutin prophylaxis for 8.4% of patients was interrupted because of adverse events. Mycobacterium avium complex (MAC) bacteremia developed in 17 (4.6%) of 371 patients receiving rifabutin prophylaxis and in 22 (11.1%) of 198 patients not receiving rifabutin prophylaxis. The mean CD4 cell count +/- SD at the diagnosis of MAC bacteremia was lower in patients receiving prophylaxis than in those not receiving prophylaxis (11.5 +/- 6.8/mm3 vs. 34.7 +/- 36/mm3, respectively; P < .01). MICs for MAC strains isolated from patients receiving prophylaxis were less than or equal to those for strains isolated from patients not receiving prophylaxis.

Thrombotic microangiopathy and cytomegalovirus disease in patients infected with human immunodeficiency virus.

Thrombotic microangiopathy (TMA) can occur during the course of human immunodeficiency virus (HIV) infection. Clinical and pathological data for 29 patients with TMA and HIV infection were recorded. In a retrospective case-control study, we analyzed the link between opportunistic infections or drug therapies and TMA. Twenty-five patients (mean CD4+ cell count +/- SD, 71.9 +/- 18.3/mm3) had renal impairment, and four had neurological dysfunction. In one-half the cases, the disease was progressive with isolated fragmentation anemia appearing several months before the clinical symptoms. The diagnosis of TMA was confirmed by histological examination of kidney biopsy specimens (18 cases). Endothelial cytomegalovirus (CMV) inclusions were associated with TMA in nine of 18 cases, whereas histological examination did not detect CMV in any control specimens (P < .001). The case-control study demonstrated a link between TMA and clinical CMV infection (odds ratio, 3.9; 95% confidence interval, 1.1-14). We conclude that TMA is a late complication of HIV infection and can be associated with systemic CMV infection in this setting.