Surgical complications and their impact on patients’ psychosocial well-being: a systematic review and meta-analysis

Objective Surgical complications may affect patients psychologically due to challenges such as prolonged recovery or long-lasting disability. Psychological distress could further delay patients’ recovery as stress delays wound healing and compromises immunity. This review investigates whether surgical complications adversely affect patients’ postoperative well-being and the duration of this impact. Methods The primary data sources were ‘PsychINFO’, ‘EMBASE’ and ‘MEDLINE’ through OvidSP (year 2000 to May 2012). The reference lists of eligible articles were also reviewed. Studies were eligible if they measured the association of complications after major surgery from 4 surgical specialties (ie, cardiac, thoracic, gastrointestinal and vascular) with adult patients’ postoperative psychosocial outcomes using validated tools or psychological assessment. 13 605 articles were identified. 2 researchers independently extracted information from the included articles on study aims, participants’ characteristics, study design, surgical procedures, surgical complications, psychosocial outcomes and findings. The studies were synthesised narratively (ie, using text). Supplementary meta-analyses of the impact of surgical complications on psychosocial outcomes were also conducted. Results 50 studies were included in the narrative synthesis. Two-thirds of the studies found that patients who suffered surgical complications had significantly worse postoperative psychosocial outcomes even after controlling for preoperative psychosocial outcomes, clinical and demographic factors. Half of the studies with significant findings reported significant adverse effects of complications on patient psychosocial outcomes at 12 months (or more) postsurgery. 3 supplementary meta-analyses were completed, 1 on anxiety (including 2 studies) and 2 on physical and mental quality of life (including 3 studies). The latter indicated statistically significantly lower physical and mental quality of life (p<0.001) for patients who suffered surgical complications. Conclusions Surgical complications appear to be a significant and often long-term predictor of patient postoperative psychosocial outcomes. The results highlight the importance of attending to patients’ psychological needs in the aftermath of surgical complications.


INTRODUCTION
Surgical complications pose significant challenges for surgical patients. Complications may vary from very minor events that can be resolved relatively quickly without the need for pharmacological treatment or other intervention, to more serious events which can be life threatening, require multiple interventions (eg, return to theatre), delay patient's discharge and may lead to multiorgan failure or even death. 1 A recent review of the literature found that postoperative complications contribute to increased mortality, length of stay and an increased level of care at discharge. 2 Other than the complications' impact on patients' postoperative recovery, they may also affect patients psychologically. They may contribute to the experience of psychological distress such as depression or anxiety due to the challenges that are inherent to them in terms of prolonged recovery or long-lasting disability (eg, severe postoperative pain, permanent

Strengths and limitations of this study
▪ This is, to our knowledge, the first systematic review of the literature assessing the impact of surgical complications on patients' psychosocial well-being. ▪ The validity of the findings is increased by the fact that only studies that used validated selfreport measures for the assessment of patients' well-being were included in the review, as well as by the use of a very comprehensive search strategy for the identification of relevant literature. ▪ Caution should be taken when interpreting these findings to other specialties as the review was limited in four surgical specialties. ▪ A limitation of this review was the very small number of studies with sufficient data for the quantitative synthesis, which did not also permit certain types of sensitivity analyses such as by surgical specialty or type of surgery. disfigurement). An early study found that patients who experienced serious adverse events after surgery reported higher levels of distress than people who had experienced serious accidents or bereavements and psychosocial adjustment worse than in patients with serious medical conditions. 3 Moreover, the authors of an interview study on patients' experiences of cardiothoracic surgery reported that a small number of patients who had a long and complicated postoperative hospital stay expressed intense feelings of hopelessness and depression. 4 Psychological distress resulting from the experience of surgical complications could further delay patients' recovery from surgery as increased levels of stress delay wound healing 5 6 and compromise immunity. [7][8][9] This review aims to critically review and synthesise the existing literature on the impact of surgical complications on adult surgical patients' psychosocial well-being and to estimate the duration of this impact. For the purpose of this review, psychosocial well-being was defined quite broadly including psychosocial outcomes of relevance to surgery such as anxiety, depression, quality of life (QoL) and post-traumatic stress. Quantitative studies which assessed the association of surgical complications with adult patients' psycho-social outcomes post-surgery were therefore reviewed. Our hypothesis was that the occurrence of surgical complications adversely affects patient psychosocial outcomes. Therefore, this systematic review aims to examine whether surgical complications impact adversely on patient psychosocial outcomes and the duration of this impact.

Search strategy
The following databases were searched through OvidSP: Terms such as patients, inpatients, outpatients, men and women were used for this facet. B. Patient psychosocial outcomes A broad definition of psychosocial outcomes was considered for the purposes of this systemic review including search terms for anxiety, depression, QoL and post-traumatic stress. 10 Two generic terms were also used, that is, well-being and emotions. The search did not include specific measures, instead it included terms for the outcomes specified above. C. Surgical complications Surgical complications were defined as any adverse event in relation to the surgical procedure including search terms for complications (eg, adverse events, untoward incidents) and terms about the surgical setting (eg, surgical, postoperative).
Each of the facets was expanded into a list of search terms truncated and combined with each other using Boolean operators, and also by mapping those to their relevant MeSH headings and subheadings in each database (through explosion of each MeSH heading). The search was restricted to titles and abstracts, and the results were limited to studies that used human participants and were written in English. The search strategies are presented in online supplementary material 1. Database searching was complemented by reviewing the reference lists of eligible articles.

Eligibility criteria
Studies were included in the review if they met the following criteria: ▸ Any quantitative study that measured the association of surgical complications with adult patients' psychosocial outcomes after surgery, either as a primary or secondary aim. Studies that measured surgical complications and psychosocial outcomes but not their association were not included as a primary analysis of reported data was beyond the scope of this review. Moreover, specific types of complications were not predefined as this review was interested in the impact of any surgical complications on patients' well-being. ▸ Psychosocial outcomes were measured with validated self-report tools or psychological assessment. ▸ Studies that reported surgical complications after cardiac, thoracic, gastrointestinal or vascular surgery, where complications are more likely to occur. 11 Studies of neuropsychological complications (eg, delirium) and studies of transplantation procedures were excluded. Conference proceedings, non-empirical data and articles that were published before the year 2000 or with the majority of their participants recruited before the year 2000 were excluded. This current approach in the selection of literature was expected to reduce bias resulting from studies of out-dated surgical practices.

Study selection
A total of 50% of the abstracts were reviewed independently by two researchers (AP and RD) and disagreements were resolved by consensus. The remaining half of the retrieved abstracts were reviewed by the primary researcher (AP) based on the consensus that was achieved for the first half. After excluding ineligible articles at abstract and title level, the remaining articles were assessed in full text. The eligibility criteria were applied again on each article. Reasons for exclusion were coded. Articles for which there was uncertainty were discussed between the primary researcher (AP), a researcher with background in psychology (RD) and a researcher with background in surgery (AA). Any disagreements were resolved by consensus.
Data extraction and quality assessment The primary researcher (AP) and a researcher with a background in surgery (AA) independently extracted data from 20 articles, which they reviewed for any disagreements. Disagreements were resolved by consensus or referral to a third senior researcher (OF). Data were extracted from the remaining articles by the primary researcher and were later checked by the second reviewer (AA). A total of 10 authors were contacted by email to provide information that was not included in the manuscripts. Three articles were excluded from the analysis because their authors did not respond to our requests for further information. Information was extracted from each article on study aims, participants' characteristics, study design, surgical procedures, surgical complications (ie, types, definitions and method of recording, where available), psychosocial outcomes (ie, scales, and time points of measurement), and the association of psychosocial outcomes with surgical complications. The latter included any reported findings on the association of surgical complications with the psychosocial outcomes, including both overall scale and subscale scores where available.
The quality of the included studies was assessed with the Newcastle Ottawa Scales (NOS). 12 The scales were modified in order to reflect the research questions of the review and to also incorporate the assessment of cross-sectional studies.

Data synthesis
The included studies were first synthesised narratively (ie, using words and text). In order to quantify the degree of the impact of surgical complications on psychosocial outcomes, quantitative procedures were also used. A meta-analysis was conducted on each extracted psychosocial outcome using Review Manager (V.5.2). 13 I 2 was used to calculate the heterogeneity present in the meta-analyses. Heterogeneity was considered low when it was below 25% and high above 50%. 14 A random-effects approach was chosen, as a degree of heterogeneity between studies should always be assumed in social sciences. 15 Where multiple assessments were conducted in one single study, only the one furthest from the participants' surgery was included in the meta-analysis.

RESULTS
In total, 18 585 articles were retrieved in total across the three databases. After removing duplicate references, a total of 13 605 papers were reviewed at abstract and title level. Nine hundred and ninety-four articles remained to be assessed in full text. A total of 51 articles (50 studies) were eligible for inclusion in the final stage of the review (see figure 1).

Study characteristics
Details of the included studies are presented in tables 1-3. A total of 28 studies were conducted in Europe, 14 in the USA, 3 in Australia, 2 in Turkey, 1 in Egypt, 1 in Japan and 1 in Taiwan. There were 29 studies in gastrointestinal,  17 in cardiothoracic [45][46][47][48][49][50][51][52][53][54][55][56][57][58][59][60][61][62] and 4 in vascular surgery. [63][64][65][66] The majority of the included studies (40 studies) assessed major procedures. The most common indications for surgery were heart conditions, followed by different types of cancer. Twenty-three studies examined the association between surgical complications and patients' well-being as a primary research aim. 17 19 28 30-38 43 47 48 50-53 55 62 64 66 The remaining examined this relationship as part of an exploration of the association of different clinical factors with patients' postoperative well-being. The majority of the studies were cohort studies. There were four case-control and 20 cross-sectional studies. The majority of the studies were prospective, including baseline measures of psychosocial outcomes.
QoL was the main reported psychosocial outcome. Three studies measured anxiety, 30 40 62 four studies measured depression 31 41 49 62 and one study measured mood states. 41 No other psychosocial outcomes were measured. The Short Form Health Survey (SF)-36 (and its associated versions, ie, SF-12, SF-20) was the most commonly used scale for the measurement of QoL. 18 61 63 The vast majority of the studies used a priori definitions of complications. For example, Bloemen et al 19 recorded only severe complications based on a grading system of surgical complications. Dasgupta et al 23 also recorded major complications which were defined as "those associated with systemic illness requiring transfer to a higher level of care or requiring relaparotomy, or complications needing interventional radiology". Others used predefined categories of complications such as infections, respiratory complications, chronic postoperative pain or perioperative myocardial infarctions. A total of 14 studies did not define or describe the complications that were recorded. The majority of the studies recorded a range of postoperative complications. Eighteen studies focused on a single category of complications (eg, anastomotic leaks, perioperative myocardial infarctions, wound complications, atrial fibrillation). Complications were mostly recorded through medical records review, clinical examinations and review of administrative databases.
Study quality varied. The scores of the included studies ranged from 2 to 8, with a mean score of 5.9. Points were deducted for the following reasons: lack of information on how complications were defined or on the methods for their recording, 16 The impact of surgical complications on patients' well-being The majority of studies (n=32) found that patients who suffered surgical complications had significantly worse postoperative psychosocial outcomes than patients with uncomplicated recovery. 16 62 65 This was the case not only after major surgical procedures but also after relatively minor operations such as hernia repairs. 18 28 30 31 43 The vast majority (n=25, 78%) were of high quality (ie, quality assessment score greater than 6 out of 8). For instance, more than half of the studies with significant findings had measured and controlled for patients' baseline psychosocial outcomes (n=18) 16  and used multivariate analyses (n=21), 16 60 62 65 suggesting that complications remained a significant independent predictor of patients' postoperative well-being even after controlling for a range of clinical and demographic factors. Psychosocial outcomes that were significantly negatively affected by surgical complications included physical, emotional and social aspects of patients' QoL as well as anxiety and depression levels (see table 4). Complications that were found to be significantly associated with worse psychosocial outcomes included both major events such as perioperative myocardial infarctions after CABG, 50 severe incontinence after internal sphincterectomy 31 or graft-related events after vascular surgery, 65 and minor complications such as wound infections after hepatic resection, 20 or new cardiac arrhythmias after CABG. 54 The complications that were significantly associated with patients' postoperative psychosocial outcomes are presented in tables 1-3.
Six studies reported a confounding association between surgical complications and patients' well-being (ie, complications were significantly associated with worse psychosocial outcomes only under certain conditions) 21 32 40 or complications were significantly associated with psychosocial outcomes at univariate but not at multivariate analysis. 49 59 64 A total of 12 studies did not find a significant association of surgical complications with postoperative psychosocial outcomes. 23 63 66 The majority of them (n=7) scored below 6 on quality assessment. For example, four studies had very small samples. 26 27 34 38 Meta-analyses A series of supplementary meta-analyses were attempted on each extracted psychosocial outcome (ie, QoL, anxiety, depression). For a meta-analysis on QoL, a synthesis of data from widely disparate assessment     9 Open Access          tools with very different composite scores (eg, social, emotional and physical) was not considered valid. For that reason, only studies that used the SF scales 67 were considered as they were the most commonly used QoL measures. Only three studies had sufficient data on the SF physical and mental QoL component scores. 28 31 45 The pooled mean differences (MD) between the two groups were statistically significant ( p<0.001), indicating lower levels of physical (MD=−3.28, CI −4.71 to −1.86) and mental (MD=−3.82, CI −4.97 to −2.67) QoL in patients who suffered complications compared with patients without complications. Two studies provided sufficient data for a meta-analysis on anxiety. 30 62 The pooled standardised MD was not significant ( p>0.05). A meta-analysis on depression was not possible as there was only one study with available data. 30 For a more detailed report of the meta-analyses, see online supplementary materials 2-4.

The duration of the impact of surgical complications on patients' well-being
Eighteen studies which reported significant associations of complications with postoperative psychosocial outcomes found a significant relationship of the presence of postoperative complications with worse psychosocial outcomes at 12 months postsurgery or later. 16 51 65 Twenty studies reported a significant association of complications with worse psychosocial outcomes at less than 12 months postsurgery. 17

DISCUSSION
This is, to our knowledge, the first systematic review of the literature investigating the impact of surgical complications on patients' psychosocial well-being. In line with our hypothesis, two-thirds of the included studies found a significant negative association between the occurrence of surgical complications and patients' postoperative well-being. The vast majority of those studies were of high quality. For instance, more than half of the studies with significant findings found that complications were an independent predictor of postoperative psychosocial outcomes after controlling for pre-existing differences on psychosocial outcomes, clinical and demographic variables.
Significant associations were reported in individual studies between surgical complications and lower scores on physical, emotional and social dimensions of the various QoL measures. A meta-analysis of three studies with sufficient QoL data collected with the SF scales suggests significant adverse effects of complications both on the physical and the mental health components. These findings are in agreement with earlier preliminary findings on the psychological burden that surgical adverse events often impose on patients. 3 4 Surgical complications were also significantly associated with higher postoperative anxiety and depression in individual studies, even though a population effect could not be shown due to the very small number of studies that measured the impact of surgical complications on anxiety and depression. Despite the fact that QoL is a useful screening outcome offering a general picture of a person's physical health and psychological state, 68 future studies on the psychosocial impact of surgical complications should also consider outcomes such as anxiety and depression as they offer a more accurate picture of a person's psychological well-being. Other relevant psychological outcomes such as post-traumatic stress, which was not measured in any of the included studies, would also be of relevance for future research in this area. It is also worth noting that strong conclusions cannot be drawn on the basis of the meta-analyses results due to the small number of studies included in them.
Complications that were found to significantly contribute to patients' low postoperative well-being ranged from severe adverse events such as anastomotic leaks after gastrointestinal surgery or perioperative myocardial infarctions after cardiac surgery to relatively minor complications such as wound infections or atrial fibrillation. It appears therefore that other than severe postoperative events, minor complications could also cause psychological distress during patients' recovery. For instance, wound complications could affect patients' satisfaction with their body image which could further compromise their QoL and psychological well-being. 69 This finding potentially implies that the severity of complications as judged by healthcare professionals does not always correspond with patients' experience of complications. Moreover, complications were negatively associated with postoperative psychosocial outcomes not only after major surgical procedures but also after relatively minor operations, 18 28 30 31 43 which suggests a potential independence of the magnitude of initial surgery with the effect of complications on patients' well-being. Further research on how complications affect patients' well-being after different types of surgery could help clarify this finding.
A number of studies also found a significant negative contribution of surgical complications to psychosocial outcomes more than 1 year postoperatively, suggesting that patients may suffer psychologically due to the experience of surgical complications for an extensive period of time after surgery. The above findings hold important implications for patients' recovery as there is growing evidence on the role of psychological stress in compromising the function of the immune system and slowing down wound healing. [7][8][9] Surgical complications are likely to further prolong patients' recovery in almost a reciprocal cycle of distress and decreased immune function. The exact relationships between surgical complications, psychological distress and speed of recovery warrant further investigation.
It is noteworthy that a smaller number of studies did not find a significant association between complications and patients' postoperative psychosocial outcomes or found significant univariate associations which were not replicated in multivariate analyses. Even in studies showing a significant impact, there will be many patients who largely maintain their psychological health and QoL in the aftermath of complications. Other than clinical factors, patients' ways of coping with stress, their appraisals of surgery and their health, as well as their perceptions of support from their loved ones and healthcare professionals could explain the conditions under which complications affect patients' well-being, as suggested by wider literature on patients' adjustment after surgical treatment. [70][71][72] The role of psychological factors as potential moderators of the psychological impact of surgical complications needs to be further explored.
Overall, the quality of the included studies was good as indicated by their relatively high-quality assessment scores and the small number of studies that scored exceptionally low. A substantial number of studies with significant findings controlled not only for patients' preoperative psychosocial outcomes but also for a variety of clinical and demographic factors confirming that surgical complications were an independent predictor of postoperative psychosocial outcomes above and beyond any pre-existing differences. The fact that the included studies used validated self-report measures for the measurement of psychosocial outcomes and the use of a very comprehensive search strategy also increase the validity of the findings.

Limitations
A few caveats should be borne in mind when interpreting the above findings. First, one-third of the studies did not define complications or did not describe the methods they used to record complications. Moreover, almost one-third of the studies did not provide information on response rates, which does not allow inferences about the representativeness of their samples.
Regarding the methodology of the systematic review, studies that were published before the year 2000 or with the majority of patients recruited before the year 2000 were excluded, albeit limiting this review to literature that was published in the last decade is expected to be more reflective of current surgical practice. It should also be noted that studies that were published past the final run of the search strategy (ie, May 2012) have not been considered. Caution should also be taken when interpreting these findings to other specialties as the clinical setting in which complications occur may affect their impact on patients' well-being. Another limitation was the very small number of studies with sufficient data for quantitative synthesis and the difficulty of synthesising data from different QoL measures, which resulted in restricting the meta-analyses on data collected only with the SF scales. The small number of studies with available data did not permit certain types of sensitivity analyses such as by surgical specialty, type of surgery (ie, minor vs major surgery) or underlying disease (eg, cancer vs other conditions), which could be significant determinants of the impact of complications on patients' well-being. Lastly, there is always the potential for publication bias where studies with significant results and big effect sizes are more easily published. [73][74][75] It is worth adding that none of the included studies were randomised controlled trials due to the non-appropriateness of this design for the research questions that this review aims to answer.

Implications of findings
The results highlight the importance of considering patients' psychological needs in the aftermath of surgical complications. Surgical and nursing staff need to be aware of the challenges of surgical complications for patients' well-being and ensure that their psychological needs are not neglected. Screening patients who suffer postoperative complications for symptoms of psychological distress could help identify those patients who need psychological support. Facilitating patients' access to psychological support during and after their hospital stay could also be of great value for patients' postoperative well-being. For example, early referral to psychological services could prevent long-term psychological distress and may also mitigate the negative effects of stress on patients' recovery. Primary care practitioners and carers need to be aware of the psychological burden that surgical complications impose on patients in order to recognise their distress in time and to provide the support that patients need.

CONCLUSIONS
This is the first systematic review of the literature on the impact of surgical complications on patients' psychosocial well-being. The findings of this review suggest that surgical complications are potentially a significant independent predictor of patients' impaired postoperative psychosocial well-being often for a very long time postsurgery. It also appears that other than major complications, relatively minor adverse events may also compromise patients' psychosocial well-being, which implies that the clinical severity of complications may not always indicate how seriously patients will be affected by them. Patients who experience surgical complications report worse levels of different aspects of QoL than patients with uncomplicated recovery, often more than a year after their operation. The ways in which complications are managed (eg, reoperation vs conservative management), the type of surgery (eg, minor vs major), the underlying disease (eg, cancer vs other conditions), psychological factors (eg, patients' perceptions of support, illness perceptions, coping strategies) or cultural influences may be key moderators of the impact of surgical complications on patients' psychosocial well-being. Future research is needed on the contribution of the above factors on the impact of surgical complications on psychological outcomes such as anxiety, depression and post-traumatic stress, as well as on how to support patients who experience a complicated postoperative recovery.
Contributors AP, OF, RD, AA and CV contributed to the conception and design of this review, and reviewed drafts of the manuscript. AP also screened all the articles retrieved by the literature searches, extracted and synthesised the data of the eligible for inclusion articles, appraised the study quality of the included articles and wrote the initial draft of this manuscript. RD screened a sample of the retrieved articles at title, abstract and full text, and AA extracted data from and scored the quality of a sample of the included articles.
Funding This work was supported by funding from the Health Foundation and the National Institute for Health Research (NIHR) Imperial Patient Safety Translational Research Centre.
Disclaimer The views expressed are those of the author(s) and not necessarily those of the funders.
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.
Data sharing statement No additional data are available.
Open Access This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) license, which permits others to distribute, remix, adapt and build upon this work, for commercial use, provided the original work is properly cited. See: http:// creativecommons.org/licenses/by/4.0/