Business & Economics

A multifactorial intervention to improve blood pressure control in co-existing diabetes and kidney disease: a feasibility randomized controlled trial

Description
A multifactorial intervention to improve blood pressure control in co-existing diabetes and kidney disease: a feasibility randomized controlled trial
Published
of 11
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
Share
Transcript
  ORIGINAL RESEARCH A multifactorial intervention to improve blood pressure controlin co-existing diabetes and kidney disease: a feasibility randomizedcontrolled trial Allison Williams, Elizabeth Manias, Rowan Walker & Alexandra Gorelik Accepted for publication 14 January 2012 Correspondence to A. Williams:e-mail: allison.williams@monash.eduAllison Williams PhD RNAssociate ProfessorSchool of Nursing & Midwifery,Monash University (Peninsula Campus),AustraliaElizabeth Manias PhD RNProfessorSchool of Health Sciences,The University of Melbourne,Victoria, AustraliaRowan Walker MDHead of NephrologyNephrology Department, Alfred Hospital,Victoria, AustraliaAlexandra Gorelik MSQASenior StatisticianClinical Epidemiology & Health ServiceEvaluation Unit,Royal Melbourne Hospital,Victoria, Australia WILLIAMS A., MANIAS E., WALKER R. & GORELIK A. (2012)WILLIAMS A., MANIAS E., WALKER R. & GORELIK A. (2012)  A multifactorialintervention to improve blood pressure control in co-existing diabetes and kidneydisease: a pilot randomized controlled trial.  Journal of Advanced Nursing   68 (11),2515–2525.  doi: 10.1111/j.1365-2648.2012.05950.x Abstract Aims.  The aim of this study was to test the feasibility and impact of an interventionconsisting of self-monitored blood pressure, medicine review, a Digital Versatile Disc,and motivational interviewing telephone calls to help people with diabetes and kidneydiseaseimprove their bloodpressurecontrolandadherencetoprescribedmedications. Background.  People with co-existing diabetes, kidney disease and hypertensionrequire multiple medications to manage their health. About 50% of people are non-adherent to their prescribed medications with non-adherence increasing in thepresence of chronic conditions. Design.  Randomized controlled trial. Methods.  Patients aged  ‡ 18 years with diabetes, chronic kidney disease and systolichypertension were recruited from nephrology and diabetes outpatients’ clinics of anAustralian metropolitan hospital between 2008–2009. Participants were randomlyallocated on a 1:1 basis to one of two groups in a randomized controlled trial: theintervention delivered over 3 months ( n  = 39) and usual care ( n  = 41), with follow-up at 3, 6 and 9 months postintervention. People collecting data and assessingoutcomes were blinded to group assignment. Results.  Seventy-five participants completed the study. The intervention wasacceptable and feasible for this cohort. There were no statistically significant dif-ferences between groups, although the mean systolic blood pressure reduction in theintervention group ( n  = 36) was   6 Æ 9 mmHg 95% CI (  13 Æ 8,   0 Æ 02) at 9 monthspostintervention. Conclusion.  The study was feasible and statistically significant differences may bedeterminable in a larger sample to overcome the variability between groups, payingattention to recommendations for further research. Trial registration.  The trial was prospectively registered with the Australian andNew Zealand Clinical Trials Register (ACTRN12607000044426). Keywords:  diabetes, hypertension, kidney disease, medication adherence, nursingintervention, self-management   2012 Blackwell Publishing Ltd  2515  JAN  JOURNAL OF ADVANCED NURSING  Introduction Diabetes and chronic kidney disease (CKD) are rapidlyescalating global health problems (Slingerland  et al.  2007,Tong & Stevenson 2007). Diabetes is the most commoncause of CKD (Atkins & Zimmet 2010), and both diseasesare frequently accompanied by hypertension (Bakris 2004).Irrespective of the cause of kidney disease, the co-existence of diabetes, CKD and hypertension leads to synergistic adverseeffects: mortality is higher, quality of life is worse and theburden on healthcare services is increased (Bakris 2004, Tong& Stevenson 2007). The global cost of diabetes is nowAU$150 billion, and will double by 2025 (Atkins 2005),whereas the total kidney failure requiring dialysis costsAU$1 Æ 1 trillion over the past decade, of which 30% isattributable to diabetes (Atkins & Zimmet 2010). Background People with co-existing diabetes, CKD and hypertensionconsult various medical practitioners in primary and special-ist settings who prescribe multiple medications for their long-term health maintenance. When multiple treatments arerequired for different health problems, medication-relatedconcerns of a person, such as adherence and side effects, aresometimes more problematic than the treatments themselves(Butler  et al.  1999, Lehane & McCarthy 2007a). Adherencecan be defined as the extent to which people follow theinstructions they are given for prescribed treatments (Haynes et al.  2008). However, approximately 50% of people do nottake their medications as prescribed with non-adherenceincreasing in the presence of chronic diseases, resulting inpoorer health outcomes (Haynes  et al.  2008). Reasons for notadhering to prescribed medications have been explored(Johnson 2002, Lehane & McCarthy 2007a), includingrecognition of the role that irrational thinking plays inmedication non-adherence (Lehane & McCarthy 2007b,Williams  et al.  2009).A systematic review of interventions to improve medica-tion adherence in people with multiple, chronic conditionsfound most studies were pharmacist-led, and focused onpreventing medication-induced morbidity rather than con-trolling chronic disease progression (Williams  et al.  2008). Amore recent review examining medication adherence inpeople > 65 years of age likely to take multiple medicationsfound simplified dosing regimens in an educational interven-tion can improve medication adherence in older people(Banning 2009). No studies were located that tested inter-ventions for improving medication adherence for people withboth diabetes and CKD, although Grant  et al.  (2003) tested apharmacist-led intervention to improve medication adherencein Type 2 diabetes and cardiovascular disease. Resultsshowed no improvement in adherence and almost half theparticipants did not complete the study due to difficultieswith follow-up.Long-term health outcomes depend on the effectiveness of the choices that people make for themselves on a daily basis(Lsrc  et al.  2001). The importance of managing people withco-existing diabetes, CKD and hypertension is critical toprevent further disease and negative health outcomes (Bakris2004). The study Aims The aim of this pilot study was to test the feasibility andimpact of a multifactorial MESMI ( Me dication  S elf- M anage-ment  I  ntervention) to improve blood pressure control andmedication adherence in adults with co-existing diabetes andCKD. We hypothesized that we would achieve a difference insystolic blood pressure of 6 mmHg between the interventionand control groups immediately postintervention at 12 weekspostbaseline and then at 3 months postintervention. Design The study was a single-centred, randomized controlled trialconducted in Melbourne, Australia, during 2008–2009.Participants were allocated on a 1:1 basis to one of twogroups: a multifactorial intervention delivered over 3 monthsand standard care. The protocol has been reported elsewhere(Williams  et al.  2010). Participants Participants were recruited from outpatient clinics at a publictertiary metropolitan hospital providing health care toapproximately 1 Æ 3 million people in Melbourne and regionalVictoria. In the planning stages of the study, we found thatmore than 1400 patients with diabetes at this hospitalhad urine microalbumin/creatinine ratios ranging from2–6020 mg/mmol indicative of kidney disease, and more than2300 patients had CKD.Patients were recruited from two weekly diabetes clinics,one weekly renal clinic and one monthly combined diabetesand nephrology clinic by a research assistant who was aRegistered Nurse. The inclusion criteria were people aged ‡ 18 years of age who comprehended English, who werementally competent (Abbreviated Mental Health test)(Hodkinson 1972), who had Type 1 or Type 2 diabetes and A. Williams  et al. 2516    2012 Blackwell Publishing Ltd  CKD estimated by a Modified Diet in Renal Diseaseglomerular filtration rate (eGFR)  > 15 ( £ 60 mL/min/ 1 Æ 73 m 2 ) or diabetic kidney disease (microalbumin/creatinineratios > 2 Æ 0 mg/mmol for men, > 3 Æ 5 mg/mmol for women),and systolic hypertension  ‡ 130 mmHg treated with pre-scribed antihypertensive medication.We had to modify our srcinal inclusion criteria to improvethe rate of recruitment because of difficulties obtainingstandard information across diabetes and nephrologypatients’ medical records. For example, some patients didnot have microalbumin/creatinine ratios taken as part of routine care, and we reduced the presence of systolichypertension for the previous clinic visit rather than theprevious two clinic visits. Patients were excluded if they livedmore than 50 km from the city centre, were pregnant or hadreceived a new diagnosis of cancer. Potential participantswere given a letter outlining the study to take home toconsider their participation before the research assistanttelephoned to confirm and organize the enrolment home visit. The intervention Participants assigned to the intervention group received amultifactorial intervention consisting of self-monitoring of blood pressure, an individualized medication review, a20-minute Digital Versatile Disc (DVD), and fortnightlymotivational interviewing follow-up telephone contact for12 weeks to support blood pressure control and optimalmedication self-management. All components of the inter-vention were delivered by an intervention nurse with renalspecialist and doctoral qualifications and trained in motiva-tional interviewing using a checklist and standing script forfidelity purposes.Participants in the intervention group were taught how totake their blood pressure, which involved being seated everymorning after breakfast and after taking their medicationsusing their non-dominant arm with an A&D Medical Pty.Ltd digital blood pressure monitor supplied for the study(Model UA-787, Saitama, Japan). Participants recorded theirdaily blood pressures for approximately 3 months in aspecific booklet issued for this purpose.The individualized medication review involved the inter-vention nurse drawing up a chart of the participant’sprescribed medications that included the generic name of themedication, what the medication was for, the dose and whento take it, and targets for which to aim. Targets included ablood pressure  £ 130 mmHg, blood glucose of  < 7 mmol/L,HbA 1C  < 7% and low-density lipoprotein cholesterol < 2 Æ 5 mmol/L as currently recommended by Harris (2008).The 20-minute DVD involved an interactive, psychosocialapproach to motivating people to take their medications,appealing to knowledge, thoughts and feelings, underpinnedby the modified Health Belief Model (Glanz  et al.  2002). Byfocusing on the attitudes and beliefs of individuals, webelieved people would be more likely to take their medica-tions as prescribed if they understood the interrelationship of their conditions, in particular hypertension, and if theirconfidence in independently managing their health wasimproved.The DVD comprised three sections: how blood pressureaffects the body; the need, benefit and safety of prescribedmedications; and tips to help people take their medications asprescribed. Targets presented in the medication review werereinforced in the DVD. Video clips of individuals withco-existing diabetes, CKD and hypertension who were notinvolved in the trial talked about their experiences of managing their medication regimens on a daily basis tofacilitate medication adherence using a psychosocialapproach.Fortnightly motivational interviewing telephone callsadapted from the guidelines of  Dilorio  et al.  (2003) com-menced 2 weeks after the intervention home visit andcontinued fortnightly until follow-up at 3-months postenrol-ment. Open-ended questions were used to prompt discussionabout the participant’s well-being, blood pressure and med-ications. Motivation and confidence in taking medicationswere assessed (Dilorio  et al.  2003), which involved obtaininga score of 1–10, with higher scores being more desirable,using this information to elicit barriers, concerns and positiveself-motivational statements. Sample size In undertaking the sample-size calculation, we examinedthe means for two independent samples. We based oursample size on Gerin  et al. ’s (2007) medicine adherence andblood pressure control (ABC) research, using the standarddeviation for systolic blood pressure of 12 Æ 1 mmHg cited inthe control group. The sample size was determined to allowthe detection of a mean ( SDSD ) difference in systolic bloodpressure of 6 mmHg between the intervention and controlgroups. The intervention group’s systolic blood pressurewas 134 mmHg, and the control group’s mean systolicblood pressure was 140 mmHg, producing an effect size of 0 Æ 495. The sample-size calculation yielded 51 participantsper group with 80% power [ a  = 0 Æ 05 (one-tailed)], includ-ing 5% attrition, totalling 108 participants in all. Randomization Following recruitment, participants were allocated codenumbers prior to enrolment and being randomized to theintervention or control group (1:1 ratio) by an off-site stat-  JAN: ORIGINAL RESEARCH  Improving blood pressure control in diabetes and kidney disease   2012 Blackwell Publishing Ltd  2517  istician. A stratified block randomization was conductedaccording to gender, age and systolic blood pressure ( £ 140 vs > 140 mmHg) recorded at recruitment. The identity of allparticipants who were enrolled and randomized to receive theintervention was kept in a locked cabinet in the chief researcher’s office. The research assistant was trained tocollect data and was blinded to group assignment. Partici-pants in the intervention group could not be blinded and wereasked to not disclose their group allocation to the researchassistant during data collection. Data collection The participants’ blood pressure was checked at each datacollection point by the research assistant in the participant’shome at enrolment, and 3, 6, 9 months postinterventionaccording to accepted guidelines (O’Brien  et al.  2003) usingan Accoson mercury desk sphygmomanometer (Accoson,London, UK) and 3M TM Littmann classic 11 S.E. stethoscope(Littman, 3M Health Care, St. Paul, MN, USA). Participantswith hypertension of  > 180 systolic mmHg on the home visitwere advised to see their primary care doctor as soon aspossible, which occurred on seven occasions.Medication adherence to all long-term prescribed medica-tions was measured by pill counts according to the processoutlined by Haynes  et al.  (1980). Insulin and over-the-counter medications such as calcium and vitamin supple-ments were not included in pill counts. At enrolment,participants were asked to keep their empty pill containersfor the duration of the study. Medication adherence wascalculated for each participant based on: P N i ¼ 1 ActualPillCount i AnticipatedPillCount i  100 N   ; where:  N   is the total number of medications.When more than 100% was calculated, adherence wastruncated at the maximum value of 100% (Haberer  et al. 2010). In addition, the four-Item Morisky MedicationAdherence Scale (Morisky  et al.  1986) was used to measureadherence to all prescribed medications with ‘yes’ and ‘no’responses. Surrogate biochemical markers of disease controltaken routinely as part of standard care, such as HbA 1C ,eGFR and serum creatinine, were collected from the patient’smedical record.Participants randomized to the control group receivedstandard care offered to patients with co-existing diabetesand CKD attending the diabetes and nephrology outpatients’clinics at the hospital. Briefly, blood pressure control was themost important aspect of standard care and care wasdependent on the patient’s individual circumstances andmorbidity. Data were collected from the control group at thesame time points as the intervention group.Medication adherence was set at  ‡ 80% as the minimumacceptable level for each prescribed medication, a commonlyaccepted cut-off point for defining a therapeutic level of medication adherence (Lee  et al.  2006). All data were enteredinto a Microsoft Office ACCESS 2003    database by asecond research assistant who was blinded to group assign-ment. An experienced biostatistician checked the integrity of the database prior to analyses.The feasibility of the study was assessed by attrition rates,the intervention group’s participation in all aspects of theintervention, and a single, open-ended question assessingtheir overall satisfaction with the intervention on its comple-tion. The impact of the multifactorial intervention wasmeasured by improved blood pressure control and medica-tion adherence. The CONSORT 2010 guidelines (Moher et al.  2010) were used to prepare this paper. Trial governance The trial was prospectively registered with the Australian andNew Zealand Clinical Trials Register before recruitment of the first participant (ACTRN12607000044426). All devia-tions and rationales from the srcinal protocol were docu-mented with the register as they occurred. The study wasconducted according to the protocol and under ethicalguidelines of the National Health and Medical ResearchCouncil of Australia (National Health and Medical ResearchCouncil 2007). Ethical considerations Prior to commencing the study, ethics approval was grantedfrom the Human Research Ethics Committees of the hospital(Project No: 2006.239) and university (Ethics ID: 0713622).At recruitment, participants were given a verbal overview of what the study entailed and a detailed written plain languagestatement to carefully consider before agreeing to participate.The enrolment home visit was organized by telephone withthe research assistant where written informed consent wasobtained. Data analysis The data were extracted and exported into  STATASTATA  10(StataCorp, College Station, TX, USA) for analysis. Thedifferences between groups were assessed using Student’s t  -test for normally distributed data (patients age, change inblood pressure, adherence rates), Wilcoxon rank-sum test forordinal data (abbreviated mental test score) and continuous A. Williams  et al. 2518    2012 Blackwell Publishing Ltd  not normally distributed data (blood test results) and Fisher’sexact test for the categorical variables. Changes in adherenceusing the Morisky scale was defined as ‘No change’,‘Improvement’ if there was a change from ‘true’ at enrolmentand ‘false’ at last follow-up, and ‘Deterioration’ if theparticipant responded ‘false’ and ‘true’ at enrolment and lastfollow-up respectively. The descriptive statistics were pre-sented as mean ( SDSD ), median (IQR) and  n  (%) respectively.The association between reduction in blood pressure andparticipant demographic, clinical (blood test results, adher-ence to medications) and other factors was assessed using theregression analysis. Statistical significance was determinedwith  P -values of  < 0 Æ 05. The analysis was performed on anintent-to-treat basis. Results Figure 1 shows the trial participant flow of participants whowere recruited between August 2008–June 2009, the majorityof whom were recruited from diabetes clinics. In-depthscrutiny of medical records revealed that a total of 12participants did not strictly meet the inclusion criteria due tonon-specific eGFR readings, for example ‘ > 60’ in theabsence of a urine microalbumin creatinine ratios to clarifythe degree of kidney disease. Twelve participants who werehypertensive in the outpatient clinic at recruitment had asystolic blood pressure of   < 130 mmHg at the enrolmenthome visit. These 24 patients therefore did not meet theinclusion criteria and were excluded from the study, of whichthe majority (62 Æ 5%) had been allocated to the interventiongroup.Immediately following enrolment, two potential partici-pants could not be contacted and four withdrew (e.g. newdiagnosis of lung cancer, objected to pill counts – ‘I pay mytaxes’). One participant who lived alone was found to beilliterate on the intervention home visit, prohibiting partic-ipation. Of these participants, 57% were in the controlgroup.Table 1 presents the baseline characteristics of the 80eligible participants enrolled in the study who had a mean ageof 67 Æ 0 (9 Æ 6) years and 56 Æ 3% were male. The majority of participants were born out of Australia ( n  = 51), mostcommonly Italy (17%) and Malta (7%). Although mostparticipants reported to follow their recommended diet, theywere obese. Three participants who were not employed wereeither on prolonged sick leave, studying full time or not Enrolment ( n   = 87)Randomized to usual care or intervention ( n   = 80) Lost to follow-up- unable to be contacted ( n   = 2) Withdrawn/illiterate ( n   = 5) Allocated to usual care ( n   = 41)Allocated to intervention ( n   = 39)Follow-up 3 months postenrolment (immediately postintervention) ( n   = 41) Follow-up 3 months postenrolment (immediately postintervention) ( n   = 38) Deceased ( n   = 1)Follow-up 6 months postenrolment (3 months postintervention) ( n   = 41) Follow-up 6 months postenrolment (3 months postintervention) ( n   = 37) Deceased ( n   = 1) Follow-up 12 months postenrolment (9 months postintervention) ( n   = 39) Withdrew ( n   = 1) Deceased ( n   = 1) Follow-up 12 months postenrolment (9 months postintervention) ( n   = 36) Incomplete data -refused BP ( n   = 1) Assessed for eligibility ( n   = 1389)Excluded ( n   = 1302) - BP/Malb/cr ratio/eGFR outside set parameters ( n   = 781) - Hypertension but not prescribed anti-hypertensives ( n   = 35) - Did not speak English ( n   = 203) - Mentally impaired ( n   = 17) - Declined participation ( n   = 123) - Eligible but missed appointment ( n   = 40) -Live >50 km CBD ( n   = 96) - New cancer diagnosis ( n   = 5) - Visual impairment ( n   = 2) Analysed ( n   = 39) Analysed ( n   = 36) Figure 1  Recruitment and participantflow.  JAN: ORIGINAL RESEARCH  Improving blood pressure control in diabetes and kidney disease   2012 Blackwell Publishing Ltd  2519
Search
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks