Using small area estimation and geographic information systems technology to target health services for the uninsured

The three articles that comprise this dissertation describe how small area estimation and geographic information systems (GIS) technologies can be integrated to provide useful information about the number of uninsured and where they are located. Comprehensive data about the numbers and characteristics of the uninsured are typically only available from surveys. Utilization and administrative data are poor proxies from which to develop this information. Those who cannot access services are unlikely to be fully captured, either by health care provider utilization data or by state and local administrative data. In the absence of direct measures, a well-developed estimation of the local uninsured count or rate can prove valuable when assessing the unmet health service needs of this population. However, the fact that these are “estimates” increases the chances that results will be rejected or, at best, treated with suspicion. The visual impact and spatial analysis capabilities afforded by geographic information systems (GIS) technology can strengthen the likelihood of acceptance of area estimates by those most likely to benefit from the information, including health planners and policy makers. ^ The first article describes how uninsured estimates are currently being performed in the Houston metropolitan region. It details the synthetic model used to calculate numbers and percentages of uninsured, and how the resulting estimates are integrated into a GIS. The second article compares the estimation method of the first article with one currently used by the Texas State Data Center to estimate numbers of uninsured for all Texas counties. Estimates are developed for census tracts in Harris County, using both models with the same data sets. The results are statistically compared. The third article describes a new, revised synthetic method that is being tested to provide uninsured estimates at sub-county levels for eight counties in the Houston metropolitan area. It is being designed to replicate the same categorical results provided by a current U.S. Census Bureau estimation method. The estimates calculated by this revised model are compared to the most recent U.S. Census Bureau estimates, using the same areas and population categories. ^

DUAL ENERGY COMPUTED TOMOGRAPHY FOR PROTON THERAPY TREATMENT PLANNING

Proton radiation therapy is gaining popularity because of the unique characteristics of its dose distribution, e.g., high dose-gradient at the distal end of the percentage-depth-dose curve (known as the Bragg peak). The high dose-gradient offers the possibility of delivering high dose to the target while still sparing critical organs distal to the target. However, the high dose-gradient is a double-edged sword: a small shift of the highly conformal high-dose area can cause the target to be substantially under-dosed or the critical organs to be substantially over-dosed. Because of that, large margins are required in treatment planning to ensure adequate dose coverage of the target, which prevents us from realizing the full potential of proton beams. Therefore, it is critical to reduce uncertainties in the proton radiation therapy. One major uncertainty in a proton treatment is the range uncertainty related to the estimation of proton stopping power ratio (SPR) distribution inside a patient. The SPR distribution inside a patient is required to account for tissue heterogeneities when calculating dose distribution inside the patient. In current clinical practice, the SPR distribution inside a patient is estimated from the patient’s treatment planning computed tomography (CT) images based on the CT number-to-SPR calibration curve. The SPR derived from a single CT number carries large uncertainties in the presence of human tissue composition variations, which is the major drawback of the current SPR estimation method. We propose to solve this problem by using dual energy CT (DECT) and hypothesize that the range uncertainty can be reduced by a factor of two from currently used value of 3.5%. A MATLAB program was developed to calculate the electron density ratio (EDR) and effective atomic number (EAN) from two CT measurements of the same object. An empirical relationship was discovered between mean excitation energies and EANs existing in human body tissues. With the MATLAB program and the empirical relationship, a DECT-based method was successfully developed to derive SPRs for human body tissues (the DECT method). The DECT method is more robust against the uncertainties in human tissues compositions than the current single-CT-based method, because the DECT method incorporated both density and elemental composition information in the SPR estimation. Furthermore, we studied practical limitations of the DECT method. We found that the accuracy of the DECT method using conventional kV-kV x-ray pair is susceptible to CT number variations, which compromises the theoretical advantage of the DECT method. Our solution to this problem is to use a different x-ray pair for the DECT. The accuracy of the DECT method using different combinations of x-ray energies, i.e., the kV-kV, kV-MV and MV-MV pair, was compared using the measured imaging uncertainties for each case. The kV-MV DECT was found to be the most robust against CT number variations. In addition, we studied how uncertainties propagate through the DECT calculation, and found general principles of selecting x-ray pairs for the DECT method to minimize its sensitivity to CT number variations. The uncertainties in SPRs estimated using the kV-MV DECT were analyzed further and compared to those using the stoichiometric method. The uncertainties in SPR estimation can be divided into five categories according to their origins: the inherent uncertainty, the DECT modeling uncertainty, the CT imaging uncertainty, the uncertainty in the mean excitation energy, and SPR variation with proton energy. Additionally, human body tissues were divided into three tissue groups – low density (lung) tissues, soft tissues and bone tissues. The uncertainties were estimated separately because their uncertainties were different under each condition. An estimate of the composite range uncertainty (2s) was determined for three tumor sites – prostate, lung, and head-and-neck, by combining the uncertainty estimates of all three tissue groups, weighted by their proportions along typical beam path for each treatment site. In conclusion, the DECT method holds theoretical advantages in estimating SPRs for human tissues over the current single-CT-based method. Using existing imaging techniques, the kV-MV DECT approach was capable of reducing the range uncertainty from the currently used value of 3.5% to 1.9%-2.3%, but it is short to reach our original goal of reducing the range uncertainty by a factor of two. The dominant source of uncertainties in the kV-MV DECT was the uncertainties in CT imaging, especially in MV CT imaging. Further reduction in beam hardening effect, the impact of scatter, out-of-field object etc. would reduce the Hounsfeld Unit variations in CT imaging. The kV-MV DECT still has the potential to reduce the range uncertainty further.

Estimation of male-to-female ratios of mutation rate in primates and rodents

Haldane (1935) developed a method for estimating the male-to-female ratio of mutation rate ($\alpha$) by using sex-linked recessive genetic disease, but in six different studies using hemophilia A data the estimates of $\alpha$ varied from 1.2 to 29.3. Direct genomic sequencing is a better approach, but it is laborious and not readily applicable to non-human organisms. To study the sex ratios of mutation rate in various mammals, I used an indirect method proposed by Miyata et al. (1987). This method takes advantage of the fact that different chromosomes segregate differently between males and females, and uses the ratios of mutation rate in sequences on different chromosomes to estimate the male-to-female ratio of mutation rate. I sequenced the last intron of ZFX and ZFY genes in 6 species of primates and 2 species of rodents; I also sequenced the partial genomic sequence of the Ube1x and Ube1y genes of mice and rats. The purposes of my study in addition to estimation of $\alpha$'s in different mammalian species, are to test the hypothesis that most mutations are replication dependent and to examine the generation-time effect on $\alpha$. The $\alpha$ value estimated from the ZFX and ZFY introns of the six primate specise is ${\sim}$6. This estimate is the same as an earlier estimate using only 4 species of primates, but the 95% confidence interval has been reduced from (2, 84) to (2, 33). The estimate of $\alpha$ in the rodents obtained from Zfx and Zfy introns is ${\sim}$1.9, and that deriving from Ube1x and Ube1y introns is ${\sim}$2. Both estimates have a 95% confidence interval from 1 to 3. These two estimates are very close to each other, but are only one-third of that of the primates, suggesting a generation-time effect on $\alpha$. An $\alpha$ of 6 in primates and 2 in rodents are close to the estimates of the male-to-female ratio of the number of germ-cell divisions per generation in humans and mice, which are 6 and 2, respectively, assuming the generation time in humans is 20 years and that in mice is 5 months. These findings suggest that errors during germ-cell DNA replication are the primary source of mutation and that $\alpha$ decreases with decreasing length of generation time. ^

AN ESTIMATION OF IN UTERO DRUG EXPOSURE BY DETERMINANTS OF FREQUENCY AND DISTRIBUTION OF MATERNAL DRUG CONSUMPTION

Data derived from 1,194 gravidas presenting at the observation unit of a city/county hospital between October 11, 1979 through December 7, 1979 were evaluated with respect to the proportion ingesting drugs during pregnancy. The mean age of the mother at the time of the interview was 22.0 years; 43.0 percent were Black; 34.0 percent Latin-American, 21.0 percent White and 2.0 percent other; mean gravida was 2.5 pregnancies; mean parity was 1.0; and mean number of previous abortions was 0.34. Completed interview data was available for 1,119 gravida, corresponding urinalyses for 997 subjects. Ninety and one-tenth percent (90.1 percent) of the subjects reported ingestion of one or more drug preparation(s) (prescription, OTC, or substances used for recreational purposes) during pregnancy with a range of 0 to 11 substances and a mean of 2.7. Dietary supplements (vitamins and minerals) were most frequently reported followed by non-narcotic analgesics. Seventy-six and one tenth percent (76.1 percent) of the population reported consumption of prescription medication, 42.5 percent reported consumption of over-the-counter medications, 45.7 percent reported consumption of a substance for recreational purposes and 4.3 percent reported illicit consumption of a substance. For selected substances, no measurable difference was found between obtaining the information from the interview method or from a urinalysis assay. ^

Slope estimation in a parametric measurement error model: A simulation study

In regression analysis, covariate measurement error occurs in many applications. The error-prone covariates are often referred to as latent variables. In this proposed study, we extended the study of Chan et al. (2008) on recovering latent slope in a simple regression model to that in a multiple regression model. We presented an approach that applied the Monte Carlo method in the Bayesian framework to the parametric regression model with the measurement error in an explanatory variable. The proposed estimator applied the conditional expectation of latent slope given the observed outcome and surrogate variables in the multiple regression models. A simulation study was presented showing that the method produces estimator that is efficient in the multiple regression model, especially when the measurement error variance of surrogate variable is large.^

Bayesian estimation of multilevel item response model with missing data

This study proposed a novel statistical method that modeled the multiple outcomes and missing data process jointly using item response theory. This method follows the "intent-to-treat" principle in clinical trials and accounts for the correlation between outcomes and missing data process. This method may provide a good solution to chronic mental disorder study. ^ The simulation study demonstrated that if the true model is the proposed model with moderate or strong correlation, ignoring the within correlation may lead to overestimate of the treatment effect and result in more type I error than specified level. Even if the within correlation is small, the performance of proposed model is as good as naïve response model. Thus, the proposed model is robust for different correlation settings if the data is generated by the proposed model.^

Estimates of hospital and physician costs of epilepsy in the United States for 1995 comparing the provider -based survey method and a patient -based medical charts and billing method

This investigation compares two different methodologies for calculating the national cost of epilepsy: provider-based survey method (PBSM) and the patient-based medical charts and billing method (PBMC&BM). The PBSM uses the National Hospital Discharge Survey (NHDS), the National Hospital Ambulatory Medical Care Survey (NHAMCS) and the National Ambulatory Medical Care Survey (NAMCS) as the sources of utilization. The PBMC&BM uses patient data, charts and billings, to determine utilization rates for specific components of hospital, physician and drug prescriptions. ^ The 1995 hospital and physician cost of epilepsy is estimated to be $722 million using the PBSM and $1,058 million using the PBMC&BM. The difference of $336 million results from $136 million difference in utilization and $200 million difference in unit cost. ^ Utilization. The utilization difference of $136 million is composed of an inpatient variation of $129 million, $100 million hospital and $29 million physician, and an ambulatory variation of $7 million. The $100 million hospital variance is attributed to inclusion of febrile seizures in the PBSM, $−79 million, and the exclusion of admissions attributed to epilepsy, $179 million. The former suggests that the diagnostic codes used in the NHDS may not properly match the current definition of epilepsy as used in the PBMC&BM. The latter suggests NHDS errors in the attribution of an admission to the principal diagnosis. ^ The $29 million variance in inpatient physician utilization is the result of different per-day-of-care physician visit rates, 1.3 for the PBMC&BM versus 1.0 for the PBSM. The absence of visit frequency measures in the NHDS affects the internal validity of the PBSM estimate and requires the investigator to make conservative assumptions. ^ The remaining ambulatory resource utilization variance is $7 million. Of this amount, $22 million is the result of an underestimate of ancillaries in the NHAMCS and NAMCS extrapolations using the patient visit weight. ^ Unit cost. The resource cost variation is $200 million, inpatient is $22 million and ambulatory is $178 million. The inpatient variation of $22 million is composed of $19 million in hospital per day rates, due to a higher cost per day in the PBMC&BM, and $3 million in physician visit rates, due to a higher cost per visit in the PBMC&BM. ^ The ambulatory cost variance is $178 million, composed of higher per-physician-visit costs of $97 million and higher per-ancillary costs of $81 million. Both are attributed to the PBMC&BM's precise identification of resource utilization that permits accurate valuation. ^ Conclusion. Both methods have specific limitations. The PBSM strengths are its sample designs that lead to nationally representative estimates and permit statistical point and confidence interval estimation for the nation for certain variables under investigation. However, the findings of this investigation suggest the internal validity of the estimates derived is questionable and important additional information required to precisely estimate the cost of an illness is absent. ^ The PBMC&BM is a superior method in identifying resources utilized in the physician encounter with the patient permitting more accurate valuation. However, the PBMC&BM does not have the statistical reliability of the PBSM; it relies on synthesized national prevalence estimates to extrapolate a national cost estimate. While precision is important, the ability to generalize to the nation may be limited due to the small number of patients that are followed. ^