Naessens, James M.

ObjectiveAssess algorithms for linking patients across de‐identified databases without compromising confidentiality.Data Sources/Study SettingHospital discharges from 11 Mayo Clinic hospitals during January 2008–September 2012 (assessment and validation data). Minnesota death certificates and hospital discharges from 2009 to 2012 for entire state (application data).Study DesignCross‐sectional assessment of sensitivity and positive predictive value (PPV) for four linking algorithms tested by identifying readmissions and posthospital mortality on the assessment data with application to statewide data.Data Collection/Extraction MethodsDe‐identified claims included patient gender, birthdate, and zip code. Assessment records were matched with institutional sources containing unique identifiers and the last four digits of Social Security number (SSNL4).Principal FindingsGender, birthdate, and five‐digit zip code identified readmissions with a sensitivity of 98.0 percent and a PPV of 97.7 percent and identified postdischarge mortality with 84.4 percent sensitivity and 98.9 percent PPV. Inclusion of SSNL4 produced nearly perfect identification of readmissions and deaths. When applied statewide, regions bordering states with unavailable hospital discharge data had lower rates.ConclusionAddition of SSNL4 to administrative data, accompanied by appropriate data use and data release policies, can enable trusted repositories to link data with nearly perfect accuracy without compromising patient confidentiality. States maintaining centralized de‐identified databases should add SSNL4 to data specifications.

Objectives: To examine multiple chronic conditions and related health care expenditures in children. Methods: Retrospective cohort study of all dependents of Mayo Clinic employees aged 0-17 on Jan 1, 2004 with continuous health benefits coverage for 4 years (N=14,727). Chronic conditions, health care utilization, and associated expenditures were obtained from medical and pharmacy claims. Results: The most prevalent chronic conditions were asthma/chronic obstructive pulmonary disease (12%), allergic rhinitis (11%), and behavior problems (9%). The most costly conditions were congenital anomalies, asthma/chronic obstructive pulmonary disease, and behavior problems ($9602, $4335, and $5378 annual cost per child, respectively). Annual health care expenditures increased substantially with the number of chronic conditions, and a small proportion of children with multiple chronic conditions accounted for a large proportion of health care costs. In addition, those with multiple chronic conditions were more likely to persist in the top 10th percentile spender group in year-to-year spending. Conclusion: Children with multiple chronic conditions accounted for a large proportion of health care expenditures. These children were also likely to persist as high spenders in the 4-year time frame. Further research into effective ways to manage the health care delivery for children with multiple chronic conditions is needed.

BACKGROUNDSurgeon and hospital factors are associated with the survival of patients treated for rectal cancer. The relative contribution of each of these factors toward determining outcomes is poorly understood.METHODSWe used data from the Surveillance, Epidemiology, and End Results–Medicare database to analyze the outcomes of patients aged 65 years and older undergoing operative treatment for nonmetastatic rectal cancer, diagnosed in the United States between 1998 and 2007. These data were linked to a registry to identify whether the treating surgeon was a board‐certified colorectal surgeon versus a noncolorectal surgeon. Hospital volume and hospital certification as a National Cancer Institute–designated Comprehensive Cancer Centers were also analyzed. The primary outcome of interest was long‐term survival.RESULTSOur data source yielded 6432 patients. Initial analysis demonstrated improved long‐term survival in patients treated by higher‐volume colorectal surgeons, higher‐volume hospitals, teaching hospitals, and National Cancer Institute (NCI)–designated Comprehensive Cancer Centers. Based on an iterative approach to modeling the interactions between these various factors, we found a robust effect of surgeon subspecialty status, hospital volume, and NCI designation. Surgeon volume was not distinctly associated with long‐term survival.CONCLUSIONSPatients treated for rectal cancer by board‐certified colorectal surgeons in centers that are higher volume and/or NCI‐designated Comprehensive Cancer Centers experience better overall survival. These differences persist after adjustment for a broad range of patient and contextual risk factors, including surgeon volume. Patients and payers can use these results to identify surgeons and hospitals where outcomes are most favorable. Cancer 2014;120:2472–2481. © 2014 American Cancer Society.

AbstractAims and objectivesThe aim of this study was to assess the impact of historical screening compliance with the effectiveness of patient reminder letters on cervical cancer screening rates.MethodUsing population‐based informatics systems, women with no cervical cancer screening in the prior 3 years were identified in two primary care clinics, Mayo Family Clinic Northeast (NE; n = 1613) and Northwest (NW; n = 1088). Patients were divided into two compliance groups: overdue/unknown screening status at study start or previously compliant. The NE Clinic sent reminder letters over a 6 months window to patients eligible for screening at the study start or during the study that were also identified as employees/dependents (E/D). There were 795 intervention (NE Clinic E/D patients) and 1906 control subjects. Using an intent‐to‐treat analysis, differences in screening rates were assessed.ResultsA higher unadjusted screening rate was observed for the E/D group than the non‐E/D group at both sites (32.7 versus 18.2% at NW, P < 0.001; 39.0 versus 14.7% at NE, P < 0.001). For the historically compliant group, unadjusted screening rates were higher for those who received letters (E/D subjects at NE) versus those who did not (E/D subjects at NW; 56.1 versus 44.5%, P = 0.01). No difference was observed between E/D subjects at NE (received letters) and NW (no letters) for the overdue/unknown group (27.4 versus 25.9%, P = 0.62). There was no difference in screening rates for non‐E/D subjects at NE versus at NW (none of whom received letters) for both the compliant (24.2 versus 30.6%, P = 0.18) and the overdue/unknown groups (11.9 versus 13.0%, P = 0.59). Multivariate logistic regression models showed a significant overall effect of E/D status (P = 0.006), compliance group (P < 0.001), and the interaction between clinic site and E/D status (P = 0.04).ConclusionAmong insured women, reminder letters appear to improve cervical cancer screening rates for those with a history of screening compliance. Reminder letters appear insufficient to motivate women if screening is overdue. Further investigation of the cohort of women overdue for screening is needed to develop interventions to successfully target this group.

Background Do Not Resuscitate (DNR) orders are intended to safeguard patients' autonomy and prevent unwanted resuscitative care. However, DNR orders may be miscommunicated between health care providers, leading to errors honoring patient wishes during cardiac arrest events. This project focused on improving accuracy of DNR ordering processes for an academic, tertiary care hospital. Intervention We describe a performance improvement process and outcomes for implementation of an inpatient electronic ordering system that included an automated, decentralized printing process for resuscitation status armbands. Specific phases of this project involved: (a) identification of common factors contributing to errors honoring patients‘ resuscitation wishes, (b) design of an electronic ordering process, (c) design and integration of a new DNR armband and (d) evaluation of the impact of changes on communication accuracy. The primary outcome was percentage of patients with incorrect designation of resuscitation status on armbands compared to the active resuscitation order in the electronic medical record. Results After implementation of an electronic ordering process we identified that 37/196 (19%) patients had an armband that did not reflect their documented wishes versus 2/103 (2%) after integration of automated armband printing into the process (p<0.001). No armband discrepancies were found after the first two weeks of post-implementation audits. Conclusions Design and implementation of an electronic ordering and armband labeling process reduced discrepancies between patient wishes and the armband labeling of the patient's desired DNR status. It is anticipated that these improvements will reduce the risk of adverse outcomes, and better align clinical processes with patient wishes.

BACKGROUND Normal Doppler hemodynamics for tricuspid prostheses have not been well characterized in a large group of patients. Therefore, we analyzed comprehensive Doppler echocardiographic examinations of 82 patients with tricuspid prostheses that were normal by clinical and two-dimensional echocardiographic examinations to establish the normal hemodynamics of various types and sizes of tricuspid prostheses. METHODS AND RESULTS The earliest complete postoperative echocardiographic study from each patient was chosen for analysis. Doppler examinations were analyzed on an off-line station from tapes or Doppler strip charts. Early velocity, atrial velocity, end-diastolic velocity, pressure half-time, and mean gradient were obtained by digitizing tricuspid velocity curves. The incidence of "physiological" tricuspid prosthetic regurgitation was noted. Ten Doppler cycles were measured for each patient, and maximal, minimal, and average measurements were recorded. The mean values +/- SD of early velocity, atrial velocity, end-diastolic velocity, mean gradient, and pressure half-time and incidence of mild prosthetic regurgitation were reported for each type of prosthesis, as were highest Doppler measurements for each valve type. Average pressure half-time was significantly lower for St Jude than for heterograft prostheses (P = .04). There were no significant differences between the valve types for mean gradient, early velocity, or incidence of prosthetic regurgitation. Increasing prosthesis size was associated with lower average pressure half-time for heterograft prostheses (P = .024). Average differences (respiratory- and cycle-length-dependent) between maximal and minimal values for 10 cardiac cycles were established for each prosthesis. CONCLUSIONS This study establishes normal ranges for Doppler hemodynamics of various tricuspid prostheses and emphasizes the importance of measuring multiple cycles for each tricuspid prosthesis, regardless of cardiac rhythm.