Sunday, June 14, 2015
One of the classic benefits touted by any clinician when describing hospice is bereavement support for the family for 12 months after the death*. What bereavement support looks like depends on the hospice agency. Some are large enough to have departments of specially trained bereavement counselors, and some have a social worker who makes it part of their role. What they offer may range from phone, email or letter correspondence to individual and group therapy sessions depending on the complexity of the case. What I have always found interesting is this is not considered part of any insurance or Medicare/Medicaid benefit for someone who dies without hospice in a hospital or nursing home. Anyways, back to hospice and bereavement support...
One of the challenges with studying what hospice does is some of the fundamental parts of it are considered standard of care without ever being tested or analyzed**. Prognostication as eligibility criteria? Yeah, that makes sense for a benefit for dying people, except the science of prognosis has not kept up with the demands of the system for ever-increasing accuracy. Interdisciplinary care required? Sure, that makes sense, but how do we know which team members really work best for each patient, we've never tested it to look for improvements. Bereavement support for families - well now we have a study which may help us understand the impact of this lauded benefit.
Ornstein et al published "Association between Hospice Use and Depressive Symptoms in Surviving Spouses" in JAMA Internal Medicine in May 2015 looked at over 1,000 widowed spouses from the Health and Retirement Study (HRS) data set. 30 percent of the patient-spouse dyads had hospice experience for more than 3 days prior to death. They compared depression scale scores between spouses who had hospice experience and those who had none (to be accurate none, or 1-2 days of hospice). 52% of widowed spouses had more depressive symptoms over time no matter the exposure to hospice. In looking at improvement in depression scale scores over time, more widowed spouses had improvement if they had hospice compared to those who did not, although it was not significantly statistic (28% vs 22% (p=.06) of all spouses; 27% vs 21% (p=.10) for spouses identified as the primary caregiver.)
The most favorable finding for hospice use to help depressive symptoms in surviving spouses was not reported in the abstract results. Of the people who were interviewed (as part of the HRS study), 38% of spouses with hospice experience had improved depressive symptoms, a statistically significant (p=.01) improvement over those who had no hospice exposure (26%). For any of you odds ratio fans, that came out to an OR of 2.15!
Most media reports gave pretty favorable headlines to this study, which I think is close enough for public knowledge and understanding. Can you use this research in your hospice 101 presentation? Sure. Can you make some nifty graphics to share online? Sure. Can we use this information to help us better structure interventions for the bereaved family? Maybe.
First, while this is a strong study in terms of numbers, matching interventions with controls, and being somewhat objective to the intervention of hospice***, the study was unable to tell what bereavement interventions occurred and therefore it is difficult to tell what interventions are helpful and which have no effect or potentially harm. So this is not a full ringing endorsement of bereavement services, but rather hospice use.
As mentioned earlier on close examination it is not truly hospice for 3 or more days vs no hospice. The true selection criteria were hospice for 3 or more days vs 0,1,or 2 days of hospice. There isn't a solid explanation in the article for this awkward choice of group selection, and the only place it is mentioned is in the footnotes of Table 4. The authors rightfully note hospice enrollment for even 1 or 2 allows for access to bereavement services, so theoretically we are not testing two different groups.
So, overall this study (despite the confounding selection choice) does help demonstrate hospice use possibly leads to better depressive symptoms, but we need to look to other studies for a more accurate depiction of the impact of bereavement services on spousal and family coping over time. This is an area of study which would benefit from a good mix of qualitative and quantitative research,
Christian Sinclair, MD, FAAHPM is a palliative care doctor at University of Kansas Medical Center, and has been a hospice medical director for 11 years. He has witnessed the good works of bereavement counselors.
*The regulations say up to a year, but nearly every hospice offers about 13 months of support to get people past the anniversary of the death.
** You know what else became standard of care without being tested or analyzed? PEG tubes in advanced cognitive impairment. It just 'made sense.' Only recently have we started to get the evidence published that it does more harm than good.
*** The depression scale was part of routine screening, not some asking you to enroll in a study of how hospice affects depressive symptoms in surviving spouses.
Ornstein KA, Aldridge MD, Garrido MM, Gorges R, Meier DE, & Kelley AS (2015). Association Between Hospice Use and Depressive Symptoms in Surviving Spouses. JAMA Internal Medicine PMID: 26009859