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.

RNA sequencing to study gene expression and single nucleotide polymorphism variation associated with citrate content in cow milk.

The technological properties of milk have significant importance for the dairy industry. Citrate, a normal constituent of milk, forms one of the main buffer systems that regulate the equilibrium between Ca(2+) and H(+) ions. Higher-than-normal citrate content is associated with poor coagulation properties of milk. To identify the genes responsible for the variation of citrate content in milk in dairy cattle, the metabolic steps involved in citrate and fatty acid synthesis pathways in ruminant mammary tissue using RNA sequencing were studied. Genetic markers that could influence milk citrate content in Holstein cows were used in a marker-trait association study to establish the relationship between 74 single nucleotide polymorphisms (SNP) in 20 candidate genes and citrate content in 250 Holstein cows. This analysis revealed 6 SNP in key metabolic pathway genes [isocitrate dehydrogenase 1 (NADP+), soluble (IDH1); pyruvate dehydrogenase (lipoamide) ß (PDHB); pyruvate kinase (PKM2); and solute carrier family 25 (mitochondrial carrier; citrate transporter), member 1 (SLC25A1)] significantly associated with increased milk citrate content. The amount of the phenotypic variation explained by the 6 SNP ranged from 10.1 to 13.7%. Also, genotype-combination analysis revealed the highest phenotypic variation was explained combining IDH1_23211, PDHB_5562, and SLC25A1_4446 genotypes. This specific genotype combination explained 21.3% of the phenotypic variation. The largest citrate associated effect was in the 3' untranslated region of the SLC25A1 gene, which is responsible for the transport of citrate across the mitochondrial inner membrane. This study provides an approach using RNA sequencing, metabolic pathway analysis, and association studies to identify genetic variation in functional target genes determining complex trait phenotypes.

Effects of supercritical fluid extraction pressure on chemical composition, microbial population, polar lipid profile, and microstructure of goat cheese.

The consumer trend for healthier food choices and preferences for low-fat products has increased the interest in low-fat cheese and nutraceutical dairy products. However, consumer preference is still for delicious food. Low- and reduced-fat cheeses are not completely accepted because of their unappealing properties compared with full-fat cheeses. The method reported here provides another option to the conventional cheese-making process to obtain lower fat cheese. Using CO(2) as a supercritical fluid offers an alternative to reduce fat in cheese after ripening, while maintaining the initial characteristics and flavor. The aim of this experiment was to evaluate the effect of pressure (10, 20, 30, and 40 × 10(6) Pa) of supercritical CO(2) on the amount of fat extracted, microbial population, polar lipid profile, and microstructure of 2 varieties of goat cheese: Majorero, a protected denomination of origin cheese from Spain, and goat Gouda-type cheese. The amount of fat was reduced 50 to 57% and 48 to 55% for Majorero and goat Gouda-type cheeses, respectively. Higher contents (on a fat basis) of sphingomyelin and phosphatidylcholine were found in Majorero cheese compared with control and goat Gouda-type cheeses. The microbial population was reduced after supercritical fluid extraction in both cheeses, and the lethality was higher as pressure increased in Majorero cheese, most noticeably on lactococcus and lactobacillus bacteria. The Gouda-type cheese did not contain any lactobacilli. Micrographs obtained from confocal laser scanning microscopy showed a more open matrix and whey pockets in the Majorero control cheese. This could explain the ease of extracting fat and reducing the microbial counts in this cheese after treatment with supercritical CO(2). Supercritical fluid extraction with CO(2) has great potential in the dairy industry and in commercial applications. The Majorero cheese obtained after the supercritical fluid extraction treatment was an excellent candidate as a low-fat goat cheese, lower in triglycerides and cholesterol but still with all the health benefits inherent in goat milk.

Generation of human cytomegalovirus-specific cytotoxic T-lymphocytes in a short-term culture.

A method to generate human cytomegalovirus (HCMV)-specific CTL (cytotoxic T lymphocytes) from human peripheral blood mononuclear cells is described. This assay is unique in comparison with other methods reported to date, because it only requires a short-term (6 days) coculture of PBM and autologous infected fibroblasts without the addition of exogenous IL-2 (interleukin-2) and nevertheless is sensitive enough to determine HCMV-specific killing in a short (6 h) 51Cr-release assay using autologous HCMV-infected fibroblasts as targets. The virus-specific killing is mediated by CTL of the CD8 phenotype and it can be inhibited by a HLA class I monoclonal antibody. The sensitivity of the assay can be significantly enhanced by pretreating the targets with interferon-gamma (IFN-gamma) prior to infection with HCMV. HCMV-specific 51Cr-release is more than doubled when the IFN-gamma pretreated targets are used. This increase is mostly due to enhanced sensitivity of the fibroblasts to killing mediated by CD8-positive CTL, but some killing can be attributed to CTL of the CD4 phenotype.

Effects of Exercise, Electrical Stimulation and Vacuum Packaging on Bacterial Counts and Tenderness of Fresh Beef Primal Cuts.

Two wholesale cuts, the silverside ( M biceps femoris ) and bolo, (outside round and clod) from 8 steers were used in this study. Four steers were artificially stressed and the right side of all carcasses was electrically stimulated. Primals were cut into 3 equal portions after 72 h post slaughter, chilling at approximately 4°C, and were vacuum packaged. No microbial differences (P>0.05) were found between primals within treatments. Primals from stressed carcasses had higher pH values (P<0.01) and psychrotrophic, lactobacillus, anaerobic and aerobic counts than from nonstressed carcasses. Lactobacilli did not dominate the microbial population. Electrical stimulation (ES) and the cuts used had an influence on shear force values (P<0.05). ES cuts were significantly more tender than controls. Results suggest that animals should be well rested before slaughter.

A new antagonist of the 5-hydroxytryptamine receptor of blood platelets.

The newly synthesized 1-desoxyfructo-alpha,beta-dehydro-4, 5-dioxotryptamine (M4) inhibited the 5-hydroxytryptamine (5-HT)-induced shape-change reaction of human and rabbit platelets and to a lesser extent the adenosine-3',5'-diphosphate-induced shape-change reaction of rabbit platelets. The 5-HT-antagonism was shown to be of the competitive type in human platelets. In the latter M4 also counteracted the 5-HT-uptake, but was virtually ineffective in releasing 5-HT. Another new derivative of desoxyfructo-5-HT (M5) was a 5-HT-agonist. It is concluded that M4 represents a new class of antagonists of 5-HT-receptors in platelets.

Shape changes induced by biologically active peptides and nerve growth factor in blood platelets of rabbits.

1 Nerve growth factor (NGF), substance P (SP) and thymopoietin all caused shape change reactions of rapid onset in rabbit platelets. NGF had the highest maximal effect, and SP the lowest EC50 (concentration causing half maximal shape change). The action of SP was reversible within 5 min, whereas that of NGF lasted for at least 1 h. A series of other peptides were inactive. 2 After preincubation of platelets with SP, a second application of SP no longer caused a shape change reaction, whereas the effect of NGF was not influenced. 3 An oxidized NGF-derivative without biological activity did not cause a shape change reaction, neither did epidermal growth factor. 4 Prostaglandin E1 (PGE1) and pretreatment of the platelets with 3% butanol, which counteract the shape changes caused by 5-hydroxytryptamine (5-HT) and adenosine 3',5'-diphosphate, also antagonized those induced by NGF and SP. Neither heparin nor methysergide, an antagonist of 5-HT-receptors, influenced the shape change induced by NGF or SP. The action of NGF was also antagonized by a specific antibody to NGF. 5 Thymopoietin, like the basic polypeptide polyornithine (mol. wt. 40,000) was not antagonized by PGE1 and butanol. Heparin, which counteracted the effect of polyornithine, did not influence that of thymopoietin. 6 In conclusion, different modes of action are involved in the shape change of blood platelets induced by polypeptides and proteins. SP and NGF may act by stimulating specific membrane receptors.

Interaction of D-LSD with blood platelets of rabbits: shape change and specific binding.

In blood platelets of rabbits, the shape change-inducing effect of D-lysergic acid diethylamide (D-LSD) has been compared with the D-LSD-binding. D-LSD, but not L-LSD, caused a shape change reaction (EC50 1.3 x 10(-9) M) which was inhibited by various 5-HT antagonists (methergoline and neuroleptic drugs), butaclamol showing marked stereospecificity. Strong inhibitors of 5 HT uptake were only weak in counteracting the D-LSD-induced shape change. Furthermore, D-[3H]LSD bound to platelets at a single saturable, high affinity, stereoselective site [Kd = (31.1 +/- 3.3) x 10(-9) M; Bmax = 28.9 +/- 2.7 fmols per 10(8) platelets]. This binding was strongly antagonized by D-LSD and methergoline and less by the hallucinogenic drugs, psilocin, bufotenin and N'N'-dimethyltryptamine. L-LSD an 5 HT, inhibitors of 5 HT uptake and neuroleptics, especially spiroperidol and butaclamol, had only a weak antagonistic effect. The latter showed stereospecificity. It is concluded that 1) the shape change reaction caused by D-LSD in platelets is mediated by the specific 5 HT-receptor, 2) the sites mediating the D-LSD-induced shape change reaction are not identical with those responsible for D-[3H]LSD-binding and both these sites are different from the 5 HT-transport sites and 3) with respect to D-LSD-binding sites, platelets and neurons are not exactly identical.

Blood platelets as models for neurons: uses and limitations.

Platelets show similarities with 5-hydroxytryptaminergic neurons with respect to (1) uptake kinetics of 5-hydroxytryptamine (5-HT) at the plasma membrane, (2) inhibitory effects of tricyclic antidepressants and neuroleptics on 5-HT uptake, (3) granular storage of 5-HT and possibly catecholamines, (4) action of drugs interfering with granular and possibly extragranular amine storage and (5) reaction of the 5-HT receptor at the plasma membrane to 5-HT agonists and antagonists. Dissimilarities include (1) the uptake of catecholamines at the plasma membrane, (2) the biosynthesis of biogenic amines (absent in platelets, present in neurons) and (3) the turnover of 5-HT (slow or absent in platelets, fast in neurons). Although the above mentioned similarities are not absolute, platelets may be considered as reasonable models for some functions of 5-hydroxytryptaminergic neurons e.g. 5-HT uptake at the plasma membrane, intracellular storage of monoamines and reactions of 5-HT receptors to drugs. In addition, the shape change reaction of platelets can probably be used to identify those basic proteins and polypeptides which cause neuronal depolarization. The significance of disturbances of the monoamine system of platelets in neuropsychiatric disorders including Parkinson's syndrome is not yet clear in all respects. Therefore, some of the current ideas about the validity of platelets as models for neurons will be briefly reviewed in this article.

Shape change of blood platelets brought about by myelin basic protein and other basic polypeptides.

Basic proteins and polypeptides (BPP) such as myelin basic protein (MBP), polyornithine (M.W. 40,000), polylysine and protamine, which are known to cause neuronal depolarization in the central nervous system, induced a shape change reaction in blood platelets of various species, including man. This reaction was not accompanied by platelet aggregation or marked alterations of 5-hydroxytryptamine release. Cyclic nucleotide levels were also unchanged. The shape change induced by polyornithine was inhibited by heparin but not by antagonists of 5HT, catecholamines or gamma-aminobutryic acid, substances which are known to have no effect on the MBP-induced neuronal depolarization. Other basic substances, e.g. low molecular weight polyornithine (M.W. 4,000), cytochrome c, spermine and spermidine, did not induce either platelet shape change or (as shown before) neuronal depolarization. It is concluded, that 1) the shape change reaction of platelets seems to be a sensitive and simple means of detecting those BPP which induce functional changes in mammalian cells and 2) the use of platelets as models for neurons can be extended to include the action of BPP on the plasma membranes.

Shape change of blood platelets induced by myelin basic protein.

Myelin basic protein (MBP) isolated from bovine spinal cord caused a marked shape change reaction of human blood platelets which was not accompanied by the release reaction and not inhibited by methysergide and spiroperidol. Only those basic proteins, including MBP, which had previously shown to exert neuronal depolarisation also induced the shape change reactions. Therefore, these findings may extend the use of platelets as neuronal models.

Uptake of 5-hydroxytryptamine in blood platelets and its inhibition by drugs: role of plasma membrane and granular storage.

The initial uptake of 3H-5-hydroxytryptamine (3H-5-HT) showed linearity for short time intervals in normal and reserpinized blood platelets of guinea-pigs, but was lower in reserpinized platelets. Teh Km values for the 3H-5-HT uptake were virtually identical in normal and reserpinized platelets, whereas Vmax was lower in the latter. Imipramine and chlorpromazine caused the same percentage inhibition of 3H-5-HT uptake in normal and reserpinized platelets; the reserpine-like compound Ro 4-1284 inhibited the uptake of 3H-5-HT in the normal, but not markedly in the reserpinized platelets. Haloperidol, prenylamine and Ro 4-9040 were more potent inhibitors in normal than in reserpinized platelets. It is concluded that (a) the Km of the initial uptake of 5-HT by platelets is probably determined by the mechanism at the plasma membrane, whereas Vmax may be codetermined by the intracellular storage capacity, (b) platelets are models for differentiating the site of action (plasma membrane or storage organelles) of drugs interfering with 5-HT uptake, and (c) neuroleptics- and reserpine-like compounds may either act selectively on the plasma membrane or on the intracellular storage organelles, or affect both of these subcellular sites.

Blood platelets as models for central 5-hydroxytryptaminergic neurons.

Blood platelets resemble 5-hydroxytryptamine (5HT) neurons of the central nervous system (CNS) with regard to uptake kinetics of 5HT at the plasma membrane and potencies of 5HT uptake inhibitors at this membrane. Furthermore, by comparing 5HT uptake in normal and reserpinized platelets the site of action of uptake inhibitors (plasma membrane, intracellular amine storage organelles) may be determined. The specific 5HT receptors of platelets whose stimulation induces a reversible shape change of platelets seem to react to drugs in a similar way as 5HT receptors of some CNS-regions such as spinal cord, cortex and possibly reticular formation. In other CNS areas e.g. those with dense 5HT innervation and the hippocampus the 5HT receptors show a reaction to drugs which is partially different from that of the platelet receptors. In other respects e.g. the synthesis and turnover of 5HT platelets do not resemble 5HT neurons. It is concluded that platelets may be used with caution as models for 5HT-neurons with regard to some aspect of 5HT-dynamics.