Search PubMed⌕ Search

Biomedical subjects

H Robbins

Publications and source records attributed to H Robbins.

At least 37 records · Page 2Linked to original sources
PubMed1982

The treatment of unicameral or solitary bone cysts by the injection of corticosteroids.

Methylprednisolone acetate injections into bone cysts are as effective a treatment as curettement and packing. More than one injection may be necessary since recurrences have been seen in half the cases treated with a single injection. Fracture of the cystic part may occur before or after injection if bony filling has not taken place. The fracture may be serious if it involves the neck of the femur. The method is safe and does not result in an unsightly scar. Solitary eosinophilic granulomata of bone also reportedly respond positively to direct injections fo methylprednisolone into the lesion by signs of healing and with relief of pain.

Adolescent↗
PubMed1981

Selecting the highest probability in binomial or multinomial trials.

Some sequential procedures are considered for selecting the binomial population with largest success probability or for selecting the multinomial outcome with highest cell probability. Procedures with and without sequential elimination of inferior populations are evaluated with respect to the expected probability of the population selected.

Journal Article↗
PubMed1980

An empirical Bayes estimation problem.

Let x be a random variable such that, given theta, x is Poisson with mean theta, while theta has an unknown prior distribution G. In many statistical problems one wants to estimate as accurately as possible the parameter E(thetax = a) for some given a = 0,1,.... If one assumes that G is a Gamma prior with unknown parameters alpha and beta, then the problem is straightforward, but the estimate may not be consistent if G is not Gamma. On the other hand, a more general empirical Bayes estimator will always be consistent but will be inefficient if in fact G is Gamma. It is shown that this dilemma can be more or less resolved for large samples by combining the two methods of estimation.

Journal Article↗
PubMed1980

Estimation and prediction for mixtures of the exponential distribution.

Let x be a random variable whose distribution is an unknown mixture of exponentials with different means theta. From a random sample x(1),...,x(n) of x values we show that E(thetax > a) can be estimated for any given a > 0. We can therefore predict the average of all future observations taken on those x values in the sample that exceed a.

Journal Article↗
PubMed1980

Sequential medical trials.

A model for sequential clinical trials is discussed. Three proposed stopping rules are studied by the Monte Carlo method for small patient horizons and mathematically for large patient horizons. They are shown to be about equally effective and asymptotically optimal from both Bayesian and frequentist points of view and are markedly superior to any fixed sample size procedure.

Journal Article↗
PubMed1980

Influencing mental health policy: the MHA approach.

The Mental Health Association was founded in 1909 as a citizen advocate organization for the mentally ill. Today it embraces more than a million members in nearly 850 chapters nationwide and conducts numerous programs designed to prevent mental illness, promote mental health, and improve mental health services. In recent years it has played an increasingly active role in influencing mental health policy at the federal level. This article focuses on the association's activities in the political arena.

Community Mental Health Services↗
PubMed1979

Local convergence theorems for adaptive stochastic approximation schemes.

For the regression model y = M(x) + epsilon, adaptive stochastic approximation schemes of the form x(n+1) = x(n) - y(n)/(nb(n)) for choosing the levels x(1),x(2),... at which y(1),y(2),... are observed converge with probability 1 to the unknown root theta of the regression function M(x). Certain local convergence theorems that relate the convergence rate of x(n) - theta to the limiting behavior of the random variables b(n) are established.

Journal Article↗
PubMed1978

Adaptive design in regression and control.

When y = M(x) + epsilon, where M may be nonlinear, adaptive regression designs of the levels x(1), x(2),... at which y(1), y(2),... are observed lead to asymptotically efficient estimates of the value theta of x for which M(theta) is equal to any desired value y(*). More importantly, these designs also make the "cost" of the observations, defined at the nth stage to be Sigma(1) (n) (x(i) - theta)(2), to be of the order of log n instead of n, an obvious advantage in medical and other applications.

Journal Article↗
PubMed1977

Strong consistency of least-squares estimates in regression models.

A general theorem on the limiting behavior of certain weighted sums of i.i.d. random variables is obtained. This theorem is then applied to prove the strong consistency of least-squares estimates in linear and nonlinear regression models with i.i.d. errors under minimal assumptions on the design and weak moment conditions on the errors.

Journal Article↗
PubMed1977

Failure of a "prophylactic" antimicrobial drug to prevent sepsis after fiberoptic bronchoscopy.

A 53-year-old patient with lymphoblastic lymphoma developed progressive interstitial pulmonary infiltrates. The patient was believed to have Pseudomonas bronchitis and was given parenteral gentamicin. Transbronchial lung biopsy performed through a fiberoptic bronchoscope was complicated by the development of Pseudomonas septicemia. The organism isolated on blood culture, like the one isolated on sputum culture, was fully susceptible to gentamicin in vitro. A prophylactic antimicrobial drug did not prevent bacteremic Pseudomonas infection in our immunosuppressed patient with pre-existent Pseudomonas bronchitis.

Bronchitis↗
PubMed1976

Maximally dependent random variables.

Let X(1),..., X(n) have an arbitrary common marginal distribution function F, and let M(n) = max(X(1),..., X(n)). It is shown that EM(n) </= m(n), where m(n) = a(n) + n[unk](an) (infinity)[1 -F(x)]dx and = F(-1)(1 - n(-1)), and that EM(n) = m(n) when X(1),..., X(n) are "maximally dependent"; i.e., P(M(n) > x) = min{1, n[1 - F(x)]} for all x. Moreover, as n --> infinity, a(n) approximately m(n) approximately m(n) (*), where m(n) (*) = EM(n) when X(1),..., X(n) are independent, provided that [1 - F(cx)]/[1 - F(x)] --> 0 as x --> infinity for every c > 1, and E(X(1) (-))(r) < infinity for some r > 0. The case in which F is standard normal is considered in detail.

Journal Article↗
PubMed1975

Sequential Estimation of p with Squared Relative Error Loss.

We consider the sequential estimation of p, the probability of success in an infinite sequence of Bernoulli trials, when the loss incurred in stopping after n trials and estimating p by some function delta(n) of the first n outcomes is taken to be [Formula: see text] The loss due to error of estimation is thus the symmetrized relative squared error, which is appropriate in applications when p may be near 0 or 1.We begin by finding a heuristic procedure for sequentially determining the sample size n which, with the usual terminal estimator of p, performs well for any fixed 0 < p < 1 as the cost per observation c --> 0. For any fixed c > 0, however, this procedure is poor as p --> 0 or 1. To remedy this defect, the uniform prior on p is introduced. The corresponding Bayes procedure is found and is shown to have a Bayes risk approximately 2pi radicalc as c --> 0.

Journal Article↗