Comparing telephone and mail responses to the CAHPS survey instrument. Consumer Assessment of Health Plans Study

OBJECTIVES

The Consumer Assessment of Health Plans (CAHPS) survey is designed to collect member experiences with getting medical care. The objective was to evaluate the comparability of answers to CAHPS questions when data are collected by mail and by telephone interview.

METHODS

Two studies comparing phone and mail responses used a pretest instrument with parallel samples drawn from Medicaid beneficiaries in California (n = 217 telephone, 97 mail) and adults with chronic conditions who had health insurance through the State of Washington (n = 98 telephone, 109 mail). A third study used a revised instrument with two parallel cross-section samples of adults covered through the State of Washington (n = 446 telephone, 609 mail). Questions covered respondents' experiences with getting medical care through their health plans.

RESULTS

In the first two tests, numerous significant differences were found in the rates at which questions that potentially did not apply to all respondents were answered: some ratings were more positive on the telephone. In the test of a revised instrument, nine of 58 comparisons differed significantly by mode. The systematic differences in response to questions that did not apply to all respondents were greatly reduced. Only one of four ratings and one of seven multi-item composite measures of quality of care were significantly different by mode.

CONCLUSION

Although further steps to reduce the remaining mode effects are needed, the data indicate that when the revised CAHPS questions are used, mode of data collection will have little effect on the key results.

The impact of having parents report about both their own and their children's experiences with health insurance plans

OBJECTIVES

The aim of this study was to determine whether parents rate their children's care differently when they also rate their own care than when they do not.

METHODS

Subjects were employees of Washington State who had been enrolled in a health plan for at least 6 months and who had at least one covered child. Subjects were randomly assigned to four study groups that were surveyed using different protocols. To assess the stability of responses over time, a follow-up telephone interview was conducted with individuals in two of the groups.

RESULTS

Parents or guardians who received both the Adult and Child Surveys were less likely to complete a survey than those who received only one survey. Responses to selected survey questions were quite stable between survey administrations. Parents who rated only their child's health care experiences generally gave more positive responses than those who also rated their own care, although few of these differences were statistically significant. This may have been due, in part, to the lower response rates in the latter group. The pairs of survey questions that ask about the adult's and child's experiences with the same aspects of care had moderate to high levels of association. The pair with the weakest association asked how clearly the doctor or nurse explained things to the adult or the child.

CONCLUSIONS

Sending both an adult and child survey to an adult could have an effect on the pattern of responses and result in lower response rates, but this might be a cost-effective way to collect reports about both adult and child health care.

Respondent selection by mail: obtaining probability samples of health plan enrollees

OBJECTIVES

This study was done to assess the feasibility of respondent selection by mail to obtain random samples of both child and adult enrollees of health plans when only subscriber contact information is available.

METHODS

The subjects were enrollees of health plans covered under the policies of employees of the State of Washington. Subscribers were eligible for inclusion in the study samples if they had been enrolled for at least 6 months and, depending on the test, had at least one child and/or a spouse enrolled under their policy. Subjects were randomly assigned to six groups: three approaches for sampling children, two approaches for sampling adults, and one test of parents' willingness to return a questionnaire about themselves. Child selection protocols involved asking respondents to complete either a child-only or a dual questionnaire, asking them to follow a decision rule to choose a sample person, and collecting data in two phases, asking respondents to return material twice.

RESULTS

Results indicated that asking subscribers to select an adult respondent by mail was not a success. At least given the procedures we used, adults did not demonstrate good compliance with the respondent selection process offered them. In contrast, parents proved willing to follow a more complicated child selection rule and to do it nearly perfectly.

CONCLUSIONS

Parents will follow a decision rule to select an eligible child, but requiring this additional respondent selection step may be associated with a slightly decreased response rate. Asking parents to return materials twice is feasible, but it is too cumbersome a procedure to be practical. It is possible to collect data about both an adult and a child using a dual instrument; however, the increase in information is tempered by a decrease in response rates.

Epilogue: Early lessons from CAHPS Demonstrations and Evaluations. Consumer Assessment of Health Plans Study

OBJECTIVES

The Consumer Assessment of Health Plans Study (CAHPS) was developed to provide an integrated set of tested, standardized surveys to obtain meaningful information from health plan enrollees and their experiences. Many organizations began to implement CAHPS in 1997. Formal evaluations of the experiences of three demonstration sites with implementing CAHPS (ie, process evaluations) and the impact of CAHPS on consumer's choices (ie, outcome evaluations) were conducted. This article reports on the early findings and feedback from our process evaluations about the sites' experiences with using CAHPS. Results are presented from the first round demonstration sites, including the lessons learned during the demonstrations. Our plans for future demonstrations and evaluations are included.

METHODS

A similar evaluation design and instruments were used across demonstration sites. The process evaluation to monitor program intervention included on-site interviews, off-site review of documents, and focus groups with consumers.

RESULTS

There are 4 early results from the CAHPS demonstrations: (1) the CAHPS survey covers topics of importance to sponsors, is of reasonable length, and can be administered quickly; (2) the report templates are being used effectively, but sponsors vary widely in their preferences for summarizing and presenting CAHPS ratings; (3) standardized or off-the-shelf products are aspects of CAHPS that sponsors value highly, while emphasizing need for further development; and (4) because surveys like CAHPS require multiple within-plan samples to make plan comparisons, they require a substantial investment and may be affordable only for large sponsors.

CONCLUSION

The first round CAHPS demonstrations highlighted the strengths of the integrated surveys and the areas for improving the products and the implementation process.

Characterization of mutations in the cystathionine beta-synthase gene in Irish patients with homocystinuria

We used single-strand conformational polymorphism and nucleotide sequencing to characterize defective cystathionine beta-synthase gene alleles in 18 independent Irish patients with homocystinuria. Six mutations were detected, three of which have been reported previously and three of which were novel. The novel mutations include T302C (L101P), C684G (N228K), and G1063C (A354P). Of the three, only T302C (L101P) was somewhat prevalent, being found in 3 of 37 independent alleles.

Squamous cell hyperplastic foci: precursors of cutaneous papillomas induced in SENCAR mice by a two-stage carcinogenesis regimen

We have conducted a series of experiments to characterize the lesions that are precursors of cutaneous papillomas in SENCAR mice initiated with 7,12-dimethylbenz(a)anthracene (DMBA) and promoted with 12-O-tetradecanoylphorbol-13-acetate (TPA). The first grossly detectable lesions at sites where papillomas subsequently developed were papules, slightly raised areas of skin ranging in diameter from 0.25 to approximately 1.5 mm. Papules were first detected in DMBA-initiated mice 21 days after the start of dosing with TPA. Of 78 DMBA/TPA-induced papules tracked during 15 weeks of TPA treatments, 68% progressed to papillomas, 9% persisted as papules, and 22% completely regressed. Histological evaluation of serial sections of 69 DMBA/TPA-induced papules revealed that they were focal hyperplastic lesions that we refer to as squamous cell hyperplastic foci (SCHF). These hyperproliferative lesions appeared to progress through two distinct stages. Stage I SCHF were characterized as regular hyperplastic foci involving the interfollicular epidermis and the outer root sheaths of 1 or more hair follicles down to the level of the sebaceous glands. Stage II SCHF were foci of irregular epithelial hyperplasia with increased fibrovascular stroma and involved from 3 to >10 hair follicles. Prominent dilated capillaries and inflammatory cell infiltrates were frequently associated with both stage I and II SCHF. Ha-ras gene codon 61 mutations were detected in 7 of 10 stage I SCHF and 13 of 14 stage II SCHF microdissected from histological sections and 7 of 7 of whole papules by mutation-specific PCR analysis. These data provide molecular evidence that SCHF are foci of initiated cells. Further study of these lesions may contribute to more fully defining the sequence of molecular and cellular changes necessary for tumorigenesis in mouse skin. SCHF may also have utility as early indicators of potential skin tumorigenicity in cancer bioassays.

Androgen responsiveness of mouse kidney beta-glucuronidase requires 5'-flanking and intragenic Gus-s sequences

Genetics studies of natural variants of the androgen response of mouse beta-glucuronidase (GUS) reveal a cis-active element closely linked to the GUS structural gene (Gus-s) that is necessary for this kidney-specific response. Results of our previous studies suggested sequences within or near an androgen-inducible deoxyribonuclease I-hypersensitive site (DH site) located in the ninth intron of Gus-s are associated with the androgen response of GUS. Using transgenic mice, we now demonstrate that at least two regions of sequence within Gus-s are involved in regulating the androgen response of GUS. The first, located within 3.8 kb of Gus-s 5'-flanking sequence, directs the response and its tissue specificity, while the second, located within a 6.4-kb fragment of Gus-s extending from the third through the ninth intron of Gus-s, protects the androgen responsiveness of the transgene from repressive influences of the insertion site.

Functional modeling of vitamin responsiveness in yeast: a common pyridoxine-responsive cystathionine beta-synthase mutation in homocystinuria

Cystathionine beta-synthase (CBS) deficiency is an autosomal recessive disorder which results in extremely elevated levels of total plasma homocysteine (tHcy) and high risk of thromboembolic events. About half of all patients diagnosed with CBS deficiency respond to pyridoxine treatment with a significant lowering of tHcy levels. We examined 12 CBS-deficient patients from 10 Norwegian families for mutations in the CBS gene and identified mutations in 18 of the 20 CBS alleles. Five of the seven patients classified as pyridoxine-responsive contain the newly identified point mutation, G797A (R266K). This point mutation is tightly linked with a previously identified 'benign' 68 bp duplication of the intron 7-exon 8 boundary within the CBS gene. We tested the effect of all of the mutations identified on human CBS function utilizing a yeast system. Five of the six mutations had a distinguishable phenotype in yeast, indicating that they were in fact pathogenic. Interestingly, the G797A allele had no phenotype when the yeast were grown in high concentrations of pyridoxine, but a severe phenotype when grown in low concentrations, thus mirroring the behavior in humans. These studies show that the G797A mutation is an important cause of pyridoxine-responsive CBS deficiency and demonstrate the utility of yeast functional assays in the analysis of human mutations.

Evidence that initiated keratinocytes clonally expand into multiple existing hair follicles during papilloma histogenesis in SENCAR mouse skin

We have previously shown that the precursors of cutaneous papillomas in SENCAR mice initiated with 7,12-dimethylbenz[a]anthracene and promoted with 12-O-tetradecanoylphorbol-13-acetate are focal hyperplastic lesions that we refer to as squamous cell hyperplastic foci (SCHF). Ha-ras gene codon 61 mutations were frequently found in SCHF, providing evidence that these lesions represent clones of initiated cells. We report here the pathogenesis of multiple hair follicle involvement in more advanced SCHF and describe the role of the hair follicle in papilloma histogenesis. Detailed histological evaluation of 83 SCHF and 25 early papillomas revealed a morphological continuum from the least developed SCHF, involving only one hair follicle, to advanced SCHF and early papillomas, which involved more than 10 hair follicles. These results provide evidence of the recruitment of additional hair follicles as SCHF progress. In advanced SCHF and early papillomas the bulk of the epithelial component in all cases consisted of several markedly hyperplastic adjacent hair follicles, whereas the involved interfollicular epidermis (IFE) was generally less hyperplastic. All of the hair follicles involved in SCHF appeared to have been preexisting, based on their pattern of spacing, that they were consistently normal appearing below the level of the sebaceous glands, and that they were in the same phase of the hair cycle as surrounding, uninvolved hair follicles. Also, no evidence of follicular neogenesis was observed in serially sectioned SCHF, and coalescence of smaller lesions was rare. To investigate whether the involvement of multiple hair follicles in SCHF was due to expansion of initiated cells into existing hair follicles or, possibly, to a paracrine mechanism, we analyzed different levels of three serially sectioned SCHF and one early papilloma for Ha-ras mutations. These analyses revealed cells with Ha-ras gene codon 61 mutations at multiple levels that involved different hair follicles. Overall, our results provide evidence that as initiated cells clonally expand, they spread across the IFE and populate the upper permanent portions of existing hair follicles. The abnormal proliferation of the infundibula of the hair follicles involved in SCHF appears to give rise to most of the epithelial component of papillomas.

Are common mutations of cystathionine beta-synthase involved in the aetiology of neural tube defects?

Mildly elevated maternal plasma homocysteine (Hcy) levels (hyperhomocysteinemia) have recently been observed in some neural tube defect (NTD) pregnancies. Plasma levels of Hcy are governed by both genetic and nutritional factors and the aetiology of NTDs is also known to have both genetic and nutritional components. We therefore examined the frequency of relatively common mutations in the enzyme cystathionine beta-synthase (CBS), which is one of the main enzymes that controls Hcy levels, in the NTD population. Neither the severely dysfunctional G307S CBS allele nor the recently reported 68 bp insertion/I278T CBS allele was observed at increased frequency in the cases relative to controls. We therefore conclude that loss of function CBS alleles do not account for a significant proportion of NTDs in Ireland.

Homocysteine and risk of premature coronary heart disease. Evidence for a common gene mutation

BACKGROUND

Plasma homocysteine levels are modulated by nutritional and genetic factors, among which is the enzyme methylenetetrahydrofolate reductase (MTHFR). A common defective (thermolabile) variant of this enzyme is causally associated with elevated plasma homocysteine, itself an independent risk factor for coronary heart disease.

METHODS AND RESULTS

To examine the hypothesis that the allele (T) that codes for the thermolabile defect increases the risk of coronary heart disease, we studied 111 patients with clinical and objective investigational evidence of coronary heart disease and 105 control subjects. The frequencies of the thermolabile defect (T) in patients and control subjects were measured, and the prevalence of elevated plasma total homocysteine according to genotype was assessed. The frequency of the defective allele was higher in patients than in control subjects with an OR of 1.6 (95% CI, 1.1 to 2.4; P = .02). The OR in the coronary heart disease group for the homozygous TT genotype was 2.9 (95% CI, 1.2 to 7.2; P = .02); 17% of patients and 7% of control subjects had the TT genotype. Plasma total homocysteine levels were significantly associated with disease status, a relationship that matched the strength of the association between disease and homozygous inheritance of the defective enzyme.

CONCLUSIONS

Homozygotes for the defective allele (T) are at increased risk of premature coronary heart disease. MTHFR, which modulates basal plasma homocysteine concentration, is folate dependent, and dietary supplementation or fortification with folic acid may reduce plasma homocysteine levels and consequent coronary risk in a significant proportion of the general population.

Androgen responsiveness of the murine beta-glucuronidase gene is associated with nuclease hypersensitivity, protein binding, and haplotype-specific sequence diversity within intron 9

The tissue specificity and genetic variability of the murine beta-glucuronidase (GUS) response to androgen provide useful markers for identifying elements which underlie this responsiveness. While GUS is expressed constitutively in all examined cell types, kidney epithelial cells uniquely exhibit a manyfold yet slow rise in GUS mRNA and enzyme levels when stimulated by androgens. Three major phenotypes of this androgen response have been described among inbred strains of mice: (i) a strong response in strains of the Gusa haplotype, (ii) a reduced response in strains of the Gusb and Gush haplotypes, and (iii) no response, as observed in Gusor mice. These response variants define a cis-active element(s) which is tightly linked to the GUS structural gene. Nuclease hypersensitivity scans of kidney chromatin within and surrounding the structural gene revealed an androgen-inducible hypersensitive site in intron 9 of the gene in Gusa but not in Gusor mice. When a radiolabeled fragment of Gusa DNA containing this hypersensitive site was incubated with kidney nuclear extracts and then subjected to gel electrophoresis, two shifted bands were observed whose levels were dramatically higher in extracts of androgen-treated than in those of untreated Gusa mice. The shifted bands reflect binding of a kidney-specific factor(s) to a 57-bp region of complex dyad symmetry in Gusa and Gusor mice which is partially deleted in Gusb and Gush mice. This binding site is located approximately 130 bp downstream of a glucocorticoid response element sequence motif which is totally deleted in [Gus]or mice. Taken together, our results suggest that the androgen responsiveness of GUS in murine kidney epithelial cells is controlled by elements within the proximal end of intron 9 of the GUS structural gene.

Accident-prone norms: implications for clinical laboratories. A methods for examining groups norms and interaction networks

Employee accidents are a costly problem for clinical laboratories. Consequently, investigation and prevention of accidents are integral parts of the clinical laboratory manager's job. There are a wide variety of traditional modes of accident research; however, they usually do not include investigations of work group norms and interaction networks. Group norms that contradict formal safety norms can arise in clinical laboratory work groups. These accident-prone norms often result in problematic procedures and activities. Examples of these accident-prone norms include: failure to use safety equipment, failure to follow safety procedures, alcohol or drug use on the job, failure to pass on information about equipment problems, and failure to report job-related accidents. These norms are frequently transmitted through interaction networks. Sociometric techniques can provide insights about the networks and can be used to develop solutions to accident-prone norms.

DNA determinants of structural and regulatory variation within the murine beta-glucuronidase gene complex

The murine beta-glucuronidase (GUS) gene complex, [Gus], encompasses the GUS structural element, Gus-s, and a set of regulatory elements which serve to modulate Gus-s expression. Three common GUS haplotypes representing virtually all inbred strains of laboratory mice have been compared with respect to GUS mRNA sequence. Results of such comparisons revealed sequence variations which target the location of one of the GUS regulatory elements to sequences within Gus-s and which account for known electrophoretic and heat stability differences among GUS allozymes of the three common GUS haplotypes.

Overlapping transcriptional units on the same strand within the murine beta-glucuronidase gene complex

We have identified and partially characterized a complex transcriptional unit within the murine beta-glucuronidase gene complex on chromosome 5. On the same strand and within the first intron of the beta-glucuronidase structural gene, Gus-s, we observe an RNA polymerase II promoter motif. That sequences within this carefully defined region can promote RNA polymerase II transcription is supported by results of in vitro transcriptional runoff assays and by expression of a linked reporter gene in both cultured cells and transgenic mice. Results of RNA blot hybridization and S1 nuclease protection studies reveal a 2.2-kilobase processed liver transcript which is initiated just downstream of the promoter motif and sharing little, if any, sequence with the 2.7-kilobase beta-glucuronidase mRNA. Both RNA species are found in liver where beta-glucuronidase is known to be expressed in all cell types. To our knowledge, this is the first description of eukaryotic mRNAs from overlapping transcription units which share the same strand yet exhibit little, if any, sequence similarity. A possible regulatory relationship between these overlapping structural genes is discussed.

Complete sequence and organization of the murine beta-glucuronidase gene

The murine beta-glucuronidase structural gene (Gus-s) has been isolated from a BALB/cJ sperm DNA bacteriophage library and its nucleotide sequence established. The gene is organized into 12 exons comprising 17.5% of the 14,009 base pair (bp) region spanning the interval between transcription initiation and the putative site of polyadenylation. A TATA box sequence, embedded within a GC-rich region, is found 28 bp upstream from the transcription initiation site. Eleven members of the B1 family and eight members of the B2 family of murine repetitive elements were identified within Gus-s and 2440 bp of flanking sequence. Other novel sequences found within Gus-s, including a (AC)19 homocopolymer tract within intron 3 and a 23 base pair complex direct repeat within intron 9, are presented and discussed.

The complete nucleotide sequence of murine beta-glucuronidase mRNA and its deduced polypeptide

The complete nucleotide sequence of murine beta-glucuronidase (GUS) mRNA has been compiled from three overlapping cloned cDNAs and a single GUS-specific genomic clone. The sequence is composed of 2455 nucleotides, exclusive of the poly(A) tail. The 5' and 3' untranslated regions contain 12 and 499 bases, respectively, with the open reading frame encoding a polypeptide of 648 amino acids (74.2 kDa), including a 22 amino acid signal sequence. The nucleotide and deduced amino acid sequences of murine GUS are compared to those published for rat and human GUS and the results are presented. Murine GUS also shares amino acid sequence identity with Escherichia coli GUS and beta-galactosidase. The complete sequences of murine GUS mRNA and its deduced polypeptide provide a basis from which to study the mechanisms responsible for the well-characterized variation in GUS expression among inbred mouse strains.

DNA sequence variation within the beta-glucuronidase gene complex among inbred strains of mice

Tightly linked to the gene that encodes murine beta-glucuronidase (GUS) are three GUS-specific regulatory elements. Together, these elements define the GUS gene complex. Specific alleles of each regulatory element are associated with a specific GUS structural allele. These associations define the three common forms (haplotypes) of the GUS gene complex, designated A, B, and H. As an initial step in defining the DNA determinants of each regulatory element and to develop DNA markers for the common haplotypes, we have identified several DNA variants by blot hybridization analysis of restricted genomic DNA using GUS-specific cDNA probes. Of 30 tested restriction endonucleases, 24 reveal DNA polymorphisms that distinguish B- and H-haplotype DNA from that of the A haplotype. Of these 24, 18 uncover a restriction fragment length polymorphism in which the polymorphic fragment of A-haplotype DNA is 200-300 bp larger than the corresponding fragment of B- or H-haplotype DNA. DNA sequence analysis of this polymorphic region reveals the presence of a short, interspersed repetitive element of the B2 family within A-haplotype DNA which is absent in DNAs of B- or H-haplotype mice. None of the DNA variations revealed by these analyses can be associated at this time with variation in the regulatory or structural properties of GUS among the common haplotypes. Nevertheless, they do provide useful haplotype-specific markers within the GUS gene complex which are of critical importance for DNA transfer experiments in transgenic mice and in cultured cells.

Outcome of patients admitted to an acute geriatric medical unit

To find out what happens to patients admitted to an acute geriatric medical unit, all admissions during 1982 were reviewed. Demographic features were compared with those of the community served, and rehabilitation, inpatient mortality and mortality in the year following discharge were assessed. Inpatients accounted for 4% of the community aged over 65, and most patients were discharged back to the community. Inpatient mortality was 25% and mortality in the year following discharge was 23%, giving a two year mortality of 42%, which was similar in all age groups. The achievement of high rehabilitation rates was tempered by the considerable mortality rates following discharge.

Genetic control of levels of murine kidney glucuronidase mRNA in response to androgen

A cis-acting genetic element, designated Gus-r, regulates the androgen-induced rates of murine glucuronidase (EC 3.2.1.31) synthesis in kidney tubule cells and is tightly linked to the glucuronidase structural gene, Gus-s. To investigate the molecular mechanism underlying this regulation, we have cloned a glucuronidase-specific cDNA sequence in plasmid pBR322. This cloned DNA has been utilized as a probe in blot hybridization analyses to determine whether the control of androgen responsiveness of kidney glucuronidase synthesis by Gus-r is exerted over the level or the translatability of glucuronidase mRNA. Three important observations emerged from these studies: (i) glucuronidase mRNA exists as a single size class of approximately 2,800 nucleotides; (ii) androgen stimulation of glucuronidase synthesis is directly related to the level of glucuronidase mRNA; and (iii) strain differences in levels of kidney glucuronidase mRNA accumulated in response to androgen are controlled by alleles of Gus-r. Thus, Gus-r regulates the androgen responsiveness of glucuronidase synthesis by controlling the amount of glucuronidase mRNA available for translation and is a cis-acting genetic element that regulates the hormonal responsiveness of a specific mRNA.

The isolation of the beta A-, beta C-, and gamma-globin genes and a presumptive embryonic globin gene from a goat DNA recombinant library

As an approach to understand how the expression of globin genes are regulated during development, clones containing globin DNA sequences were selected from a recombinant library of goat genomic DNA. The type of globin gene present in each of the recombinants was determined by cross-hybridization to the DNA of mouse alpha- and beta-globin cDNA-containing plasmids. Of 11 clones isolated, eight hybridized specifically to the DNA of the mouse beta-globin plasmid, while one clone hybridized only to the DNA of the alpha globin plasmid. The location of each globin sequence within its DNA insert was determined by a combination of restriction enzyme mapping and Southern transfer-hybridizations. Selected fragments were sequenced; comparisons of the amino acids coded for by these regions with those of the goat globins identified clones carrying beta A-, beta C-, and gamma-globin genes. Another recombinant coded for amino acid sequences resembling, but not identical with, the known goat globins, and was identified tentatively as containing an embryonic or epsilon-gene. Detailed analysis of the clone containing the beta C gene and an overlapping clone revealed that three other beta-like sequences are located 6, 12, and 21 kilobases on the 5'-side of the beta C gene. The globin sequence of the locus nearest to the beta C gene has an altered translation termination codon and, if transcribed and translated, would give a globin chain seven amino acids longer than the normal goat beta C-globin. In addition, the sequence following this termination codon is very AT-rich, unlike that of other globin genes. The recombinants described contain extensive regions of DNA surrounding the globin genes, making them useful for identifying regulatory sequences as well as determining the sequence organization of the goat globin genes.

Residue organic mixtures from drinking water show in vitro mutagenic and transforming activity

Indications of possible health effects of residue organics in drinking water have been sought using short-term tests of mutagenic and transforming activity. Ten percent or less of the total organic material in drinking water has been identified; the remainder is believed to include thousands of unknown nonvolatile compounds. Residual organics were concentrated from drinking water from representative U.S. cities by reverse osmosis followed by liquid-liquid extraction [yielding the reverse osmosis concentrate-organic extract (ROC-OE) fraction] and sorption-desorption on XAD-2 resin. Samples of these residue organics were provided by the Environmental Protection Agency for bioassay. They were examined for mutagenic activity by using Salmonella tester strains (primarily TA98 and TA100) and for transforming activity by using mouse fibroblasts (BALB/3T3 clone 1-13). City-specific patterns of dose-dependent bacterial mutagenesis and of bacterial toxicity were observed for these samples and for subfractions generated by sequential extractions with hexane, ethyl ether, and acetone. Mutagenic effects were essentially independent of a microsome activation system prepared from liver of Aroclor 1254-induced rats. On the basis of strain-specific effects in mutagenesis and differential distributions of mutagenic activity during liquid-liquid extraction, at least some of the active compounds are thought to be acidic, frameshift mutagens. The ROC-OE fraction of a New Orleans sample transformed BALB/3T3 cells in replicate experiments. By comparison with the bacterial mutagenesis data, cell transformation is a relatively sensitive method for detecting possible mutagenic and carcinogenic activity in this sample. The appropriateness of these systems for the assay of complex mixtures and the degree to which reverse osmosis concentrates contain the unaltered organic compounds in the original samples are discussed.