Comparing disease specific catastrophic cost estimates using longitudinal and cross-sectional designs: The example of tuberculosis

BACKGROUND

There has been an increasing interest in assessing disease-specific catastrophic costs incurred by affected households as part of economic evaluations and to inform joint social/health policies for vulnerable groups. Although the longitudinal study design is the gold standard for estimating disease-specific household costs, many assessments are implemented with a cross-sectional design for pragmatic reasons. We aimed at identifying the potential biases of a cross-sectional design for estimating household cost, using the example of tuberculosis (TB), and exploring optimal approaches for sampling and interpolating cross-sectional cost data to estimate household costs.

METHODS

Data on patient incurred costs, household income and coping strategies were collected from TB patients in Negros Occidental and Cebu in the Philippines between November 2018 and October 2020. The data collection tools were developed by adapting WHO Tuberculosis Patient Cost Surveys: A Handbook into a longitudinal study design. TB-specific catastrophic cost estimates were compared between longitudinal and simulated cross-sectional designs using different random samples from different times points in treatment (intensive and continuation phases).

RESULTS

A total of 530 adult TB patients were enrolled upon TB diagnosis in this study. Using the longitudinal design, the catastrophic cost estimate for TB-affected households was 69 % using the output approach. The catastrophic cost estimates with the simulated cross-sectional design were affected by the reduction and recovery in household income during the episode of TB care and ranged from 40 to 55 %.

CONCLUSION

Using longitudinally collected costs incurred by TB-affected households, we illustrated the potential limitations and implications of estimating household costs using a cross-sectional design. Not capturing changes in household income at multiple time points during the episode of the disease and estimating from inappropriate samples may result in biases that underestimates catastrophic cost.

Costs incurred by people with co-morbid tuberculosis and diabetes and their households in the Philippines

OBJECTIVE

Diabetes is a risk factor for TB mortality and relapse. The Philippines has a high TB incidence with co-morbid diabetes. This study assessed the pre- and post-TB diagnosis costs incurred by people with TB and diabetes (TB-DM) and their households in the Philippines.

METHODS

Longitudinal data was collected for costs, income, and coping mechanisms of TB-affected households in Negros Occidental and Cebu, the Philippines. Data collection was conducted four times during TB treatment. The data collection tools were developed by adapting WHO's cross-sectional questionnaire in the Tuberculosis Patient Cost Surveys: A Handbook into a longitudinal study design. Demographic and clinical characteristics, self-reported household income, number of facility visits, patient costs, the proportion of TB-affected households facing catastrophic costs due to TB (>20% of annual household income before TB), coping mechanisms, and social support received were compared by diabetes status at the time of TB diagnosis.

RESULTS

530 people with TB were enrolled in this study, and 144 (27.2%) had TB-DM based on diabetes testing at the time of TB diagnosis. 75.4% of people with TB-DM were more than 45 years old compared to 50.3% of people with TB-only (p<0.001). People with TB-DM had more frequent visits for TB treatment (120 vs 87 visits, p = 0.054) as well as for total visits for TB-DM treatment (129 vs 88 visits, p = 0.010) compared to those with TB-only. There was no significant difference in the proportion of TB-affected households facing catastrophic costs between those with TB-DM (76.3%) and those with TB-only (68.7%, p = 0.691).

CONCLUSION

People with TB-DM in the Philippines face extensive health service use. However, this does not translate into substantial differences in the incidence of catastrophic cost. Further study is required to understand the incidence of catastrophic costs due to diabetes-only in the Philippines.

Experiences, challenges and looking to the future in a clinical tuberculosis cohort in the time of COVID-19 in the Philippines

A cohort study of Filipino tuberculosis patients is currently undergoing data collection amidst the coronavirus disease 2019 pandemic. In this article we present the current experiences, challenges and obstacles of our team during this period as we attempt to fulfil our roles and responsibilities in Metro Manila, Cebu and Negros Occidental in the Philippines. Each site had different lockdown restrictions and experienced problems to different degrees. The underlying themes were similar, covering the supply chain, mobility, communication, physical and mental health and disruption of health services due to reallocation of staff. While we maximized the use of mobile devices, logistical challenges remained. Institutional support for the field teams, creative problem solving and resilience are required to adapt in a rapidly changing environment.

Diabetes mellitus in patients with liver cirrhosis

In spite of the high prevalence of diabetes mellitus (DM) in patients with liver cirrhosis (LC) few studies have focused on the clinical implications of this association. We investigated the clinical and pancreatic-endocrine features of 34 patients who developed DM after LC (Group I). Results were compared with 34 carefully matched patients with only Type II DM (Group II). A standard meal test was performed in 26 patients with normal renal function from each group to assess beta-cell function. Group I patients, less frequently had retinopathy (14.7% vs. 45.5%, P < 0.05) and a family history of diabetes (23.5% vs. 58.8%, P < 0.01). Group I patients also showed signs of enhanced insulin resistance, reflected by higher insulin dose requirements in insulin-treated patients (0.87 +/- 0.10 vs. 0.62 +/- 0.05 IU/kg/day, P < 0.01) and increased basal C-peptide values (0.88 +/- 0.06 vs. 0.68 +/- 0.07 pmol/l, P < 0.05, respectively) than those in Group II. These results suggest that several clinical features, probably related to the hepatopathy, define DM occurring in patients with LC.

Symptomatic and hormonal hypoglycaemic responses to human and porcine insulin in patients with type I diabetes mellitus

In recent years there has been great concern that human insulin (HI) may induce fewer hypoglycaemic warning symptoms than porcine insulin (PI). We addressed this issue in eight patients aged 25.6 +/- 3.3 (SEM) years with Type I (insulin-dependent) diabetes mellitus of 15.1 +/- 3.7 years duration who complained that hypoglycaemia unawareness had appeared after transferring from PI to HI. Acute induction of hypoglycaemia was induced on two occasions with semisynthetic HI and purified PI under double-blind conditions. Blood glucose was first clamped for 2 h at 4.4-6.7 mmol l-1 with an intravenous infusion of HI or PI at 50 mU kg-1 h-1 and 20% glucose at a variable rate. Thereafter, insulin infusion alone was maintained for 100 minutes. Heart rate, arterial pressure, reflex times, autonomic and neuroglycopenic signs and symptoms were assessed every 10 min. Arterialized venous blood samples were taken to measure blood glucose every 10 min and catecholamines, insulin, glucagon, growth hormone, and cortisol every 20 min. Autonomic symptoms first appeared at a plasma glucose level of 2.92 +/- 0.21 mmol l-1 with HI vs 2.92 +/- 0.48 mmol l-1 with PI (NS). There were no significant differences between the two studies concerning any of the above mentioned clinical parameters or the counterregulatory hormone responses. A differential effect of insulin species on the ability to perceive hypoglycaemia in patients who ascribed diminished perception of hypoglycaemia to the use of HI was thus not observed.

Characterization of the clinical features of five families with hereditary primary cutaneous lichen amyloidosis and multiple endocrine neoplasia type 2

The hereditary conditions of primary cutaneous lichen amyloidosis and multiple endocrine neoplasia type 2 (MEN 2) are rare clinical entities. The initial reports of two families in which the two conditions coincided have led to the identification of at least eight additional families with this clinical syndrome. In this report we describe the clinical features in five of these eight families. The salient feature in these five families is the presence of unilateral (46%) or bilateral (64%) pruritic and lichenoid skin lesions located over the upper portion of the back. Family members describe these skin lesions as intermittently intensely pruritic leading to scratching and excoriation of the upper back region. The presence of MEN 2 has been documented in 97% of family members with this skin lesion, the one exception being a child who is at risk for development of MEN 2A in whom the diagnosis has not yet been made. Of family members who have MEN 2A, 27% do not have an identifiable skin lesion, although the skin lesion developed in one patient two years after a curative thyroidectomy for medullary thyroid carcinoma (MTC). Four of the five families have members with pheochromocytoma; one with five affected members has only MTC. The finding of this clinical syndrome in geographically diverse portions of the world and the lack of overlap with MEN 2A without the skin lesion suggest it is a distinct clinical variant of MEN 2A.

Cyclosporine's effect on insulin secretion in patients with kidney transplants

Adverse effects of cyclosporine on glucose metabolism have been reported in patients with kidney transplantation, but steroids were present in all the immunosuppressive schedules evaluated. We have studied endocrine pancreatic function (OGTT measuring plasma insulin [IRI] and c-peptide [CP]) in 3 groups of patients, matched for age and body-mass index, 16-66 months after functioning Ktx. Seven patients were treated with CsA monotherapy (group I), 7 patients with CsA plus prednisone (group II), and 6 patients with azathioprine plus prednisone (group III). Seven healthy subjects formed the control group. OGTT was normal in all patients, except one in group II (impaired glucose tolerance). There were no significant differences between the 4 groups concerning fasting blood glucose and area under the glucose curve, as well as basal insulin levels, peak insulin response to glucose, and area under the insulin curve. Basal CP, peak CP response to glucose, and area under CP curve were lower in CG than in the 3 groups of patients. Basal CP in group II (4.4 +/- 2.2 ng/ml) was higher than in group I (2.8 +/- 0.6 ng/ml, P less than 0.05). Glucose/IRI molar ratio in group II (5.7 +/- 1.4, P greater than 0.05) was lower than in group I (7.3 +/- 1.8) and CG (8.0 +/- 2.1, P less than 0.025). Our results suggest that CsA at normal dosage has no clinically important effect on beta-cell function. The indirect evidence of insulin resistance observed in patients treated with CsA plus prednisone is ascribable to corticosteroid treatment.

Primary localized cutaneous amyloidosis and familial medullary thyroid carcinoma

We have studied a family with an autosomal dominant inheritance of primary localized cutaneous amyloidosis (PLCA) and familial medullary thyroid carcinoma (MTC). Ten family members were screened for multiple endocrine neoplasia (MEN) 2; five were found to have MTC and two had C-cell hyperplasia. None had evidence of phaeochromocytoma or parathyroid abnormalities. Five of these seven patients presented characteristic interscapular hyperpigmented lesions, showing dermal amyloid deposits in two of the four patients in which a biopsy was performed. The data are analysed in the light of two recent reports of MEN 2A associated with identical lesions. We conclude that PLCA should be sought in MTC patients, even if no other endocrinopathies are present. This may be informative of the familial character of MTC in index cases and also of the tumour gene status in family members who are being screened.

Auditory function in young patients with type 1 diabetes mellitus

Comprehension of the effect of diabetes mellitus on auditory function has been hindered by the fact that previous studies have evaluated hearing function in heterogeneous groups of patients with diabetes mellitus, thus giving conflicting results. We have performed audiometric studies in 46 consecutive patients. 13 with newly diagnosed type 1 diabetes mellitus (group 1) and 33 with type 1 diabetes mellitus of more than 3 years of duration (group 2), of 14 to 40 years of age. The results were compared to an age-matched control group. Pure-tone auditory thresholds were significantly higher in all frequencies 250-8,000 Hz in both groups when compared to the control subjects. Ten patients, all of which belonged to group 2, had auditory thresholds above 30 dB in at least one frequency, showing a conversational hearing loss that ranged between 11 and 44%. However, none of them referred subjective hypoacusia. Univariate analysis revealed significant associations between auditory thresholds and age, duration of disease as well as retinopathy, but not with neuropathy, HbA1c or hypoglycaemic episodes. Only age and duration of disease independently correlated with an auditory threshold using multiple regression. We conclude that type 1 diabetes mellitus can cause mild sensorineural hearing impairment which correlates with age and duration of disease.

Asymptomatic electronystagmographic abnormalities in patients with type I diabetes mellitus

Only a few studies have investigated vestibular function in diabetes mellitus (DM), showing contradictory results. We have performed an electronystagmographic (ENG) evaluation of 46 individuals with type I DM and 37 healthy controls. No patient reported subjective vestibular symptoms. Duration of caloric-induced nystagmus (DN) was significantly lower (2.1 +/- 0.7 vs. 2.6 +/- 0.4 min, p less than 0.01), and central nystagmus frequency of caloric response also nonsignificantly tended to be decreased (37.4 +/- 16.5 vs. 41.7 +/- 12.7 beats/30s, p = 0.21) in DM patients, as compared to controls. The latter comparison achieved significance after exclusion of newly diagnosed diabetic patients (33.4 +/- 16.1 vs. 41.6 +/- 12.7 beats/s, p less than 0.05). Depressed caloric reactions were seen in 21.8% of patients. DN was lower in patients with microalbuminuria and retinopathy, but this was not observed after exclusion of newly diagnosed diabetic patients, all of whom had normal ENG responses and no chronic diabetic complications. The existence of a lower DN and central nystagmus frequency should be borne in mind when interpreting ENG tracings in patients with long duration type I diabetes mellitus.

Effects of thromboxane synthesis inhibitor triflusal on renal hemodynamics in microalbuminuric diabetic patients

Triflusal (2-acetoxy-4-trifluormethylbenzoic acid) is a platelet-antiaggregant drug that selectively inhibits thromboxane synthesis with little effect on prostacyclin production. In this study, we evaluated the effect of 5-day administration of 900 mg/day triflusal on glomerular filtration rate (GFR), renal plasma flow (RPF), urinary albumin excretion (UAE), thromboxane B2 (TXB2), 6-ketoprostaglandin F1 alpha (PGF1 alpha), and PGE2 in nine normotensive insulin-dependent diabetic patients with UAE between 30 and 103 micrograms/min. Plasma TXB2 and plasma renin activity (PRA) were also determined. After administration of triflusal, we observed a reduction in microalbuminuria (59 +/- 25 vs. 33 +/- 22 micrograms/min, P less than 0.01), an increase in RPF (648 +/- 119 vs. 722 +/- 134 ml.min-1 x 1.73 m-2, P less than 0.01), and a reduction in filtration fraction (0.24 +/- 0.04 vs. 0.20 +/- 0.03, P less than 0.01). Triflusal produced a significant reduction in both plasma TXB2 (130 +/- 39 vs. 52 +/- 32 pg/ml, P less than 0.02) and urine TXB2 (523 +/- 249 vs. 312 +/- 11 pg/min, P less than 0.02), without changes in PRA and UAE of 6-keto-PGF1 alpha and PGE2. Metabolic control and arterial blood pressure did not change during the study. These results suggest that platelet-antiaggregant therapy can reduce microalbuminuria in diabetic patients. This effect could be mediated by a reduction in the transglomerular hydraulic pressure through a vasodilation of efferent arterioles secondary to renal thromboxane synthesis inhibition.

Fulminant liver failure and pancreatitis associated with the use of sulfamethoxazole-trimethoprim

We report the case of a 26-yr-old patient with fulminant liver failure and acute hemorrhagic pancreatitis secondary to the use of trimethoprim-sulfamethoxazole (Bactrim DS). Our patient presented with skin rash and decreased C3 and C4 levels, which we believed was due to a hypersensitivity reaction secondary to the sulfonamide component (sulfamethoxazole). To our knowledge, this is the first case reported in which sulfamethoxazole-trimethoprim has been implicated as a cause of fulminant liver failure and acute hemorrhagic pancreatitis simultaneously, and emphasizes the need of discontinuing this medication as soon as there is evidence of liver and pancreatic dysfunction.

Fulminant idiopathic pseudomembranous colitis

Pseudomembranous colitis is characterized by inflammatory plaques and pseudomembranes on the colonic mucosa. The disorder most commonly occurs after the use of antibiotics, which allow overgrowth of Clostridium difficile, a spore-forming, gram-positive rod that produces a toxin. Overgrowth of Staphylococcus aureus can also produce pseudomembranous colitis. In rare cases, pseudomembranous colitis is not associated with antibiotic use. When C. difficile is present, vancomycin or metronidazole usually produces a prompt response. In idiopathic cases, surgery may be required.

Cystadenocarcinoma and carcinoembryonic antigen (CEA)

A 40-year-old woman with back pain, an abdominal and a serum CEA level of 200 ng/ml was found to have at laparotomy, a large but resectable mucinous retroperitoneal mass. Pathologic diagnosis revealed cystadenocarcinoma, probably of pancreatic origin. It is remarkable that the cyst fluid contained a CEA level 100,000-fold normal. Serum CEA fell to normal levels in the postoperative period. This suggests that the CEA was of cyst epithelial origin.