Baseline depressive symptoms, personal control, and concern moderate the effects of preoperative psychological interventions: the randomized controlled PSY-HEART trial

Cardiovascular diseases are associated with a high risk of mortality, disability, a decreased quality of life, and increased costs for the healthcare system (Murray & Lopez, 2013; Virani et al., 2020). Coronary artery bypass grafting (CABG) is an established treatment option for patients with advanced coronary artery disease that has been thoroughly studied over several decades (Hawkes et al., 2006). However, it is still unknown why a substantial number of patients faces problems in the recovery process and does not benefit as much from the surgery as surgeons would predict (Blumenthal et al., 2003; Burg et al., 2003; Hawkes & Mortensen, 2006; Hawkes et al., 2006; Salzmann, Euteneuer, et al., 2020; Salzmann, Salzmann-Djufri, et al., 2020). Patients’ recovery after surgery is not explained by medical factors alone; recovery seems to be a multidimensional phenomenon in which physical, psychological, and social factors play important roles as well (Auer et al., 2016; Hawkes et al., 2006; Sadeghi et al., 2017). Growing evidence suggests the importance of psychological preparation for improving post-surgery physical outcomes and psychological outcomes (i.e., quality of life, disability, pain, morbidity, length of hospital stay) (Auer et al., 2017; Levett & Grimmett, 2019; Salzmann, Euteneuer, et al., 2020; Salzmann, Salzmann-Djufri, et al., 2020; Wynter-Blyth & Moorthy, 2017). A better understanding of whether and when psychological interventions affect specific outcomes may help design even more powerful interventions and make better predictions of which patients will benefit from which psychological intervention. More specifically, in this article, we assessed whether baseline depressive symptoms moderated the intervention effects on depressive symptoms, whether baseline anxiety levels moderated the intervention effects on anxiety levels, and whether baseline illness beliefs moderated the intervention effects on illness beliefs (i.e., whether baseline control beliefs moderated the intervention effects on control beliefs).

The Common Sense Model (CSM) aims to explain how people react to a perceived threat. It describes that patients have illness beliefs or perceptions about the experience of their illness. According to the CSM, patients experience surgery as a threat, triggering cognitive and emotional processes (Leventhal & Cameron, 1987; Leventhal et al., 1992). These processes affect illness behavior by triggering strategies to cope with the threat. The cognitive response to a health threat consists of a person's subjective illness beliefs and expectations about the identity (symptoms), the timeline (how long it will continue), the consequences (results of the symptoms), and the perceived controllability (personal and treatment control—possibility of recovery through my acting or medical treatment) of an illness (Kidd et al., 2016; Leventhal et al., 2001).

Illness beliefs have been consistently shown to be related to short-term and long-term heart surgery outcomes, the recovery process, and behavioral change and therefore may help to explain why surgery is more effective for some patients than for other patients (Broadbent et al., 2009; Juergens et al., 2010; Parfeni et al., 2013; Petrie et al., 1996, 2002; Poole et al., 2015; Salzmann, Euteneuer, et al., 2020; Salzmann, Salzmann-Djufri, et al., 2020; Weinman et al., 2000). In particular, personal control describes a person’s belief that he/she is confident to execute a specific behavior and that this behavior will affect one’s health (Laferton et al., 2017). Personal control is a behavior outcome expectation that describes how much patients are convinced that they can recover from or control the disease by their own action (Lau & Hartman, 1983; Leventhal et al., 2001). An association was found between higher personal control beliefs and better quality of life/well-being, lower depression and anxiety in CABG patients (Broadbent et al., 2015; Gallagher & McKinley, 2009; Kidd et al., 2016; Petrie et al., 2002, 2012). Changing illness beliefs has enhanced health outcomes in several studies with cardiac patients (Davies et al., 2008; Juergens et al., 2010; Keogh et al., 2011; Petrie et al., 2002, 2012). Research may benefit from focusing more on patient beliefs and expectations, especially about personal control, in exploring the recovery process. Higher scores of preoperative perceived control have been shown to predict postoperative quality of life and lower levels of depression in CABG patients (Kidd et al., 2016). Nonetheless, little is known about the question of which preoperative psychological intervention can influence what kind of illness beliefs and who will benefit from such an intervention specifically (Kidd et al., 2016).

Besides the cognitive responses to a perceived health threat, the CSM highlights the importance of emotional factors in coping with a disease, e.g., illness beliefs such as concern or emotions (Leventhal & Cameron, 1987; Leventhal et al., 1992). Other emotional factors such as depression or anxiety are also highly relevant in cardiac surgery patients: Depression is highly prevalent in patients undergoing CABG (Blumenthal et al., 2003; Head et al., 2013; Poole et al., 2014, 2017; Tully et al., 2009; Young et al., 2019). 20–40% of CABG surgery patients are affected by depression (Blumenthal et al., 2003; Connerney et al., 2001; Young et al., 2019). Depressed patients undergoing CABG surgery report a lower health-related quality of life, have a higher postoperative rate of depression, a higher risk of rehospitalization and death, and stay longer in hospital after the surgical procedure independent from medical factors compared to non-depressed patients (Auer et al., 2017; Blumenthal et al., 2003; Connerney et al., 2001; Contrada et al., 2004; Mallik et al., 2005; Morone et al., 2010; Oxlad et al., 2006; Rollman et al., 2009; Rumsfeld et al., 2004; Timberlake et al., 1997). Dunkel et al. (2011) suggest that patients with higher depression levels might benefit most from additional psychological intervention. Similarly, preoperative anxiety seems to be associated with a negative postoperative course, yet fewer is known about this relationship, especially in CABG patients (Arthur et al., 2000; Heilmann et al., 2016; Lamarche et al., 1998; Székely et al., 2007). Since preoperative anxiety and depressive symptoms seem to be important predictors of the postoperative recovery process, psychological interventions targeting these symptoms could improve postoperative physical and psychological outcomes (i.e., such as depression or anxiety). Heilmann et al. (2016) reported that a preoperative intervention reduced the preoperative and postoperative state anxiety compared to a control group. However, it is mainly unknown how specific interventions can be tailored to the individual needs of CABG patients to reduce patients anxiety and depression levels.

The PSY-HEART trial indicated that receiving a preoperative psychological intervention aiming to optimize patients’ expectations (EXPECT) led to reduced illness-related disability as the primary outcome (Rief et al., 2017). Several positive effects on secondary outcomes were also found: For instance, the EXPECT intervention indicated increased physical and mental quality of life 6 months after CABG surgery and fewer days of hospitalization in comparison to standard medical care (SMC) only (for further information, see Auer et al., 2017; Rief et al., 2017). For depression (another secondary outcome), a non-significant trend was found in favor of EXPECT and SUPPORT (an attention control group receiving the same amount of time and attention by the psychologist but without working specifically on expectations) compared to SMC only, when assessing baseline and follow-up scores 6 months after surgery. However, it is still unclear who benefitted the most from the preoperative psychological interventions in the PSY-HEART trial and when these interventions seemed to work regarding patients’ depressive and anxiety symptoms as well as illness-beliefs. A meta-analysis of Bower et al. (2013) indicated that patients who had higher levels of depression at baseline showed greater treatment effects than patients with lower levels of depression at baseline. For anxiety, baseline anxiety was the most frequently examined moderator of the effectiveness of psychological and psychoeducational interventions for anxiety in a meta-analysis (Moreno-Peral et al., 2020). Since patients’ depressive and anxiety symptoms and patients’ illness beliefs (i.e., perceived personal control or concern) are considered important outcome predictors, especially in heart surgery patients (Salzmann, Euteneuer, et al., 2020; Salzmann, Salzmann-Djufri, et al., 2020), a more thorough understanding of the (psychological) intervention effects over time regarding these psychological factors is crucial. To better understand how, when and for whom the preoperative psychological interventions (EXPECT: optimizing expectation group; SUPPORT: emotional support/attention control group) seemed to improve depressive and anxiety symptoms, and illness beliefs in the PSY-HEART trial (Rief et al., 2017), this secondary exploratory analysis examined whether (i) baseline scores of depressive symptoms moderated the effects of the preoperative psychological interventions on depressive symptoms in heart surgery patients 1 day before surgery, 4 to 6 days after surgery and 6 months after surgery, whether (ii) baseline anxiety symptom scores moderated the effects of the preoperative psychological interventions on anxiety symptoms in heart surgery patients 1 day before surgery, 4 to 6 days after surgery and 6 months after surgery, and whether (iii) baseline scores of the illness beliefs (i.e. perceived personal control or concern) moderated the effects of the preoperative psychological interventions on illness beliefs (i.e. perceived personal control or concern) in heart surgery patients 1 day before surgery, 4 to 6 days after surgery and 6 months after surgery.

This article aimed to explore whether baseline levels of patients’ depressive/anxiety symptoms and illness beliefs moderated the effects of preoperative psychological interventions (EXPECT/SUPPORT) on these constructs in heart surgery patients to develop a better understanding of whether and when psychological interventions may have an effect on these important outcomes in heart surgery patients.

Baseline levels of depressive symptoms, personal control, and concern seemed to moderate the intervention effects on depressive symptoms, personal control, and concern. Especially for patients with high baseline scores of depressive symptoms, both preoperative psychological interventions led to reduced levels of depressive symptoms 6 months after surgery compared to the control group SMC. Considering the Minimally Clinically Important Difference (MCID) of HADS scores for patients with cardiovascular disease that is ≥ 1.7 (Lemay et al., 2019), at least all discovered statistically significant effects of HADS were also clinically meaningful (≥ 1.763), except the discovered effect of SUPPORT and SMC for patients with average baseline levels of depressive symptoms at T3 (1.526). Also, some effects were found for illness beliefs, especially for personal control and concern: For a low personal control baseline-score the EXPECT-intervention led to increased personal control 1 day before surgery compared to SUPPORT and SMC. For patients with low baseline concern scores, the EXPECT-group indicated increased levels of concern 1 day before surgery compared to SMC. These results implicate that both preoperative psychological interventions may be especially relevant for patients with higher baseline depressive symptoms. Patients with lower baseline perceived personal control may benefit from the EXPECT-intervention.

As mentioned above, depression is a risk factor for many negative physical and psychological outcomes such as the higher risk of major adverse cardiac events, mortality, higher levels of medical complications, longer hospital stays, and lower quality of life after CABG surgery (AbuRuz, 2019; Auer et al., 2017; Blumenthal et al., 2003; Burg et al., 2003; Flaherty et al., 2017; McKenzie et al., 2010). Our study also indicated an association between depressive symptoms 6 months after surgery and postoperative complications leading to rehospitalization. Due to the importance of depression for psychological and physical outcomes, some studies tried to reduce depression in patients undergoing CABG surgery (Heilmann et al., 2016; McKenzie et al., 2010; Rollman, Belnap, LeMenager, Mazumdar, Houck et al., 2009; Rollman, Belnap, LeMenager, Mazumdar, Schulberg, & Reynolds III, 2009). The Bypassing the Blues (BtB) trial showed that CABG-patients who were depressed after their surgery profited from a collaborative care program (Rollman & Belnap, 2011). In the BtB-trial, patients participated after their surgery and had at least 1–28 contacts (median = 10) with the care manager in a period of eight months in the intervention groups. Compared to this trial, our results suggest that even a brief preoperative psychological intervention (five contacts) may be an effective way to decrease long-term levels of depressive symptoms and may improve heart surgery outcomes in patients with higher levels of depressive symptoms before surgery.

However, in our trial even most of the patients with ‘high’ baseline depressive symptoms did not reach the cut-off criterion for a clinical diagnosis of a depressive disorder (depression score ≥ 8) (Bjelland et al., 2002). Therefore, future research should examine how high a depressive burden has to be for letting patients benefit from a preoperative psychological intervention before undergoing surgery. Our results may not only be of scientific interest.

They may have important clinical implications: CABG-patients with significant depressive symptoms have almost twice the risk of having a cardiac event in the first 6 months after CABG-procedure (Scheier et al., 1999). In our trial, an association between depressive symptoms 6 months after surgery and complications leading to rehospitalization was found. Therefore, participation in a preoperative psychological intervention and ongoing support after the surgery (e.g., using booster calls) should be offered to these patients to improve psychological and physical outcomes. Given that further studies can replicate these findings, in line with the findings from Rollman and Belnap (2011), CABG patients should be screened for depressive symptoms before undergoing surgery. If depressed CABG-patients have a higher risk for adverse events such as cardiac events or mortality (Blumenthal et al., 2003; Scheier et al., 1999), patients with high levels of depressive symptoms who benefit from the individual surgery preparation would get the necessary assistance, while patients with lower levels (who may not benefit from the preoperative psychological interventions according to our results) would not need to take time for the intervention before and after surgery. Such a tailored treatment would provide every patient with the most profit and the fastest recovery possible. Also, the clinic and the healthcare system save capacities for the patients who need support and costs for less helpful interventions, or hospital stays (if the patients' recovery is faster and the time patients stay in hospital is shorter) (Auer et al., 2017; Oxlad et al., 2006).

Future reimbursement models will most probably focus more on outcome parameters including quality of life (value based medicine), than on diagnoses and procedures (DRG systematics). Accordingly, interventions that can positively impact the short and long-term outcomes may be worthwhile from medical and economic perspectives. The results for depression also showed that patients with high or average levels of baseline depressive symptoms receiving the SUPPORT intervention benefitted 1 week and 6 months after the CABG surgery. Patients with low levels of baseline depressive symptoms receiving the SUPPORT intervention benefitted 1 day before the CABG surgery. Thus, additional psychosocial support seems to be helpful.

No significant differences between EXPECT (optimizing expectations) and SUPPORT (focusing on emotional support) were found for depressive symptoms. Against the background of the CSM, this may be explained by the fact that depressive symptoms can be characterized as an emotional and interpersonal challenge and a cognitive alteration; patients suffering from depression indicate dysfunctional cognitions and maladaptive information processing (Beck, 1979). Both interventions included at least one placebo mechanism (optimizing expectations, empathic relationship between patient and provider) (Schedlowski et al., 2015) targeting these factors. This may have led to no significant differences between the intervention groups for depressive symptoms. Further research should focus on the question, which kind of intervention is the most helpful for patients with depressive symptoms to further elucidate crucial mechanisms.

Personal control as one of the illness beliefs seems to play an important part in the effects of preoperative psychological interventions. Previous studies indicated that personal control was most amenable to change compared to other illness belief dimensions (Broadbent et al., 2015; Laferton et al., 2016). In our trial, we found that receiving EXPECT or SUPPORT led to increased levels of personal control 1 day before surgery. Patients seemed to be more convinced to recover from or control their heart disease by their action after both preoperative psychological interventions than SMC patients. This could be due to the fact that patients pay attention to themselves and their heart disease in both interventions and therefore get the idea that their behavior influences their recovery. Furthermore, it is conceivable that patients in the SUPPORT group had the opportunity to self-reflect and may thus have come to the conclusion that their personal behavior may be part of their recovery. Since patients were not blind about their psychological intervention, receiving any kind of preoperative psychological intervention (compared to receiving only SMC) could have contributed to increased personal control levels in both psychological interventions.

The analyses indicated that patients with low perceived personal control levels at baseline benefitted from EXPECT in the short term compared to both other groups. Higher perceived personal control is associated with a higher quality of life and lower levels of depressive symptoms after CABG surgery (Kidd et al., 2016). Therefore, the increase of perceived personal control in our study might help avoid or reduce depressive symptoms in patients and hereby reduce physical outcomes as explained above. By now, only a short-term effect was found. Future studies should examine whether a sustained effect of increased personal control after surgery would have additional positive effects on long-term outcomes. Therefore, it should be examined, if more booster sessions can maintain the increase in perceived personal control for patients with lower baseline levels of personal control.

For patients with average and high baseline levels of personal control, the SUPPORT intervention led to increased levels of personal control 1 day before surgery. This finding may lead to the assumption that validation and emotional support may increase patients’ perceived controllability as someone strengthens the patients' confidence and trust in their thoughts and preparations.

Regarding patients’ level of concern, a short-term effect was observed for patients with low scores of baseline concern. In the EXPECT group, an increase of concern was observed after the intervention 1 day before surgery. By focusing on psychoeducational aspects in this group, it is not surprising that patients’ worries increased short-dated. This result may explain why no effect was found for anxiety. By focusing on realistic expectations, patients also discussed topics that may have been perceived as concerning.

Some limitations need to be considered when interpreting the results of the study. Patients were only included in the study when they could appear in the study hospital a few days before the planned surgery date. Therefore, only patients with enough interest, time, and the possibility to drive to the hospital (even if some lived far away) were included. These facts may limit the generalizability of the findings. When getting informed, patients received the information that three treatment groups are included in the study and that two of them will receive additional conversations. It is possible that the expectation of receiving “just the standard of care” or “something special, additional” may have affected the outcomes. Focusing on depressive symptoms, most of the patients did not reach the cut-off criterium (depression score ≥ 8) (Bjelland et al., 2002). Therefore, the patients in the trial were not depressed on a high level. Further, due to the explorative character of the analyses conducted, our findings should be interpreted with caution. No correction for multiple testing has been done. Multi-centered confirmatory trials including more patients are needed to confirm the findings, generalize from one study site to the general healthcare systems, investigate further clinical outcome variables, and gain more knowledge, who would benefit from which intervention. It would be important to replicate the findings with a larger sample focusing on physical and psychological symptoms such as hospital stay, mortality, rehospitalization, depression, anxiety, illness beliefs and their associations.

In conclusion, our findings indicate that some patients may benefit from preoperative interventions while others will not. This study indicated that brief preoperative psychological interventions might improve critical psychological outcomes such as depressive symptoms or personal control in some heart surgery patients, but not in all patients. It further indicated that this may especially apply to specific subgroups of patients (i.e., high baseline depressive symptoms, low baseline personal control). Patients’ psychological status at baseline may moderate the effectiveness of psychological interventions. The second important finding is that assessing baseline levels is essential to offer tailored psychological interventions to improve long-term heart surgery outcomes. Gathering patients’ psychological status before undergoing heart surgery and providing psychological interventions if they are indicated (e.g., for patients with high scores of depressive symptoms or low levels of perceived control) would be beneficial. More studies are needed to examine which patients may benefit from what kind of preoperative psychological intervention at which timepoint and why.